When looking at all this, one must look at the entire picture from start to finish, not just one little piece.
Power, audio, video, and radio frequency cables all share certain properties and have certain differences. Power and audio are relatively low frequencies while video and radio are high. Different cable types are used for each. While each may be able to share a common type of cable, it is not always recommended. Different cable types have different bandwidth properties. Super expensive mega cables do not make the signal better. These may transferr it easier, but then again, a thicker conductor may also.
Cables obey the laws of physics. They have resistance, capacitance (resistance to voltage change), and inductance (resistance to current change). Longer cables will have higher resistance, capacitance, and inductance thus degrading signal quality more than shorter ones. Thicker cables carry more amperes "easier". The thicker wire has less resistance. Thinner cables can carry high amperes, but they tend to get too hot because of higher resistance and may become a fire hazzard and even melt. As frequency increases: resistance increases, inductance decreases, capacitance decreases.
Each metal in a wire has a particular resistance per foot measured in ohms. Copper's resistance/foot is less than gold. Silver's resistance/foot is less than copper's. Silver isn't used as often because of its higher price and its tendency oxidize more than copper. Gold is typically only used in platings since it is very expensive, has a higher resistance than copper, and has a very high resistance to oxidation. Aluminum is sometimes used as wire, but it has a slightly higher resistance/foot than gold. Steel is rarely used as communications wire but is often used in connectors. Steel has a significantly higher resistance/foot than aluminum.
Expensive oxygen free copper (OFC) cables generally have no place in the home theater setup. OFC cables are generally used in sterile industrial environments because of their chemical purity rather than their slightly better conductivity. By nature regular copper cables are nearly oxygen free already. OFC cables do have slightly less corrosion problems and their conductivity is about 1% higher than regular copper cables. If one desires better conductivity, it is far more cost effective and easier to get a slightly larger gauge of wire (that will out perform any oxygen removal to begin with). Oxygen is used in the refining process to help improve purity (which is a good thing). Impurities in copper can increase its resistance. High temperature operation of copper with high oxygen content can make it brittle after extended use. Not all OFC's conduct better (especially if phosphorus is used to remove the remaining oxygen). It is pointless to spend excessive money on OFC cables when the circuit board traces are regular copper. Moisture, smog, and ozone (room air ionizers) cause more copper corrosion than oxygen in the copper. When the copper wire gets insulated, many volatile compounds are introduced that are worse than oyxgen (also problems with long term outgassing).
Some may think that if a thicker cable is better, then a super monstrous cable, pipe, or solid bar would be superb. "Bigger is better" is subjected to the laws of diminishing returns. Overly thick cables may also suffer from increased capacitance and inductance thus negating the increased conductivity.
Some high end cables mix smaller and larger copper strands to form the overall cable. The manufacturers think that the smaller strands will carry the higher frequencies better than the larger strands. Generally speaking, this is pointless. Higher frequencies have higher resistance (impedance), thinner strands have higher resistance, and thinner strands have less surface area for the skin effect. All of these together would seem to indicate that higher frequencies preferring thicker strands. It is also just as likely that the higher ampere low frequencies may somehow try to push themselves onto the thinner strands and push off the higher frequencies (isn't physics almost magical?). Keep in mind that all the strands are connected at both the starting and ending points (there is no special pre-separation). What matters most in signal transmission is the total cross sectional area (which makes up the gauge). A cable with fat insulation won't do better than a thicker copper cable with thinner insulation. Cables with mixed strand size are often smaller gauge wise than their equivalent counter parts.
The Skin Effect. DC power and low frequencies will essentially use the entire wire if viewed as a cross section. Higher frequencies tend to migrate to the outside of the wire. Audio frequencies at the very edge of human hearing start to do this but no where near what the sales and marketing companies would have you believe. For audio, cable resistance at that frequency is far more of an issue than available surface area for the high frequency signal to travel. Video and radio frequencies are different and require a shielded coax wire instead of a plain copper wire.
Litz wire is a special type of wire where each strand is thin coated in an insulating material and carefully braided together in a special pattern. Litz wire minimizes the skin effect and impedance at higher frequencies by increasing the overall surface area compared to the solid wire equivalent. Litz wire at audio frequencies is mostly pointless. Litz wire can generally go up to a few megahertz in frequency.
Solid vs stranded wire. Solid wire is cheaper to make than stranded. It can be used in situations where flexibility is not required. Solid wire has less high frequency loss due to the skin effect than the regular (non-Litz) stranded variety. Stranded is easier to bend and has some tolerance if one or more strands break. Stranded wire may reduce the proximity effect (current crowding) at radio frequencies (which may be more of a problem than the skin effect). In general, at radio frequencies a coax cable should be used instead of a regular wire.
Connecting and connectors. All wires are connected to something in one way or another. Not all connectors are equal. Better connectors will have a high surface area connection to both the wire and what the wire is being connected to. A lower surface area connection can effectively increase the impedance between the wire and the device connected to. Connectors are often made of steel for strength. Steel has a much higher resistance than gold or aluminum. Since most connectors are short and small, this has very little effect at lower frequencies. At radio frequencies, this can get complicated. Radio frequencies often reflect at splices and connectors. The cables must be impedance matched to help minimize this. Audio cables need like a mile or so before they start running into impedance reflection problems.
Long cable runs can act like an antenna. Noise increases with length. Twisting internal signal wires and forming a balanced cable with some extra hardware can cancel this out, but at the price of extra hardware for transmission and receiving. Balanced runs only really work at low and audio frequencies.
Shielding on cables helps to intercept errant noise and dump it to ground. Shielded cables used in video and radio are typically the coax variety. More shielding is better to the point of diminishing returns. Layered shielding is often better than a single shield. Braided shields tend to protect the inner wire against low frequency noise by provide a higher conductivity to ground than foil. Braided shields are also easier to attach to connectors than foil. Foil shields are often better at protecting against high frequency noise. Braided and foil shields are often combined in the better cables to take advantage of both their properties. Insulation between the layered shields is pointless since they are electrically joined at one or both ends of the cable. Double shielding may be used in particulary noisy environments. In this case, the outer shield is connected to an earth ground and the inner shield is used as the signal path return.
Shielding also serves to help hold the transmitted signal inside the cable preventing the wire from becoming a transmission antenna (think water pipe). This is mainly an issue at radio frequencies and higher.
Shields can either be connected at one end only or both ends of the cable. If both ends, it won't matter if the cable is plugged in backwards and the shield is usually used as the signal path return. If only one end is connected, it may make a difference, but that's highly dependant on the setup and grounding situation. In this case there will be a signal wire and ground return wire protected by the shield.
Noise sources should obviously be minimized at the source. Do not run signal level cables parallel to power level cables. Stay away from high noise sources like flourescent lights. If this cannot be avoided add as much distance as is possible between them (inverse square laws) and try crossing signal and power cables at right angles.
The square wave test is often used to test for the possible bandwidth of a cable. The edges of a square wave will start rounding out at the frequency and distance limits of a cable. Reflections in a cable can be seen as another ghosted square wave on top of the primary one.
When doing engineering like this, one must focus on the ENTIRE picture and not just one little piece...unless that one little piece is causing loads of problems and is at the top of the list. Generally, when something is going wrong, there is usually more than one thing causing that problem.
Subjective claims about cable quality are poor arguments and are generally not tolerated. "Smooth", "more open", "soft", and "vivid" may be accurate descriptions for true changes but are not scientific descriptions. ABX type tests should be used to determine the extent of these. Any changes should be able to be described by the scientific guidelines outlined above.
There are generally 3 main possibilities when someone swaps out a cable and claims an improvement. 1) The cable was pathetic, damaged, or the wrong one used to begin with. 2) The equipment the cable pluged into had very poor inputs and/or outputs and upgrading the cable to something a little more friendly made a difference. 3) The moron spending an absurd amount of money on the cable will naturally claim it made a positive difference to avoid looking foolish. This is usually the case.
When trouble shooting or upgrading, the cables are often the least of the worries in the setup (so long as they all connect as they should). There are far many other problems that should be addressed first, and the entire signal chain from start to finish should be examined.
Why spend an absurd amount on a cable when the other cables used by the TV/cable/satellite company are nowhere near the specs claimed by the wannabe super cable? Especially when the broadcasts are poorly balanced to begin with? (A digital cable card and a good VGA monitor can be an interesting peek into their "lackings".)
Why spend an absurd amount on a power cable when it would be plugged into a $10 power strip which is plugged into a $0.50 wall outlet that has $10 worth of 14 gauge gauge house wire going into a $10 circuit breaker? People who claim a huge difference here need to get their head examined.
Why spend $60 on an A/V cable that connects to a $0.30 connector on the back of a DVD player? Even if the external connector just happened to be good, the internal connections from that connector may not be. Television connectors on the other side of that cable are notorious for being low quality.
A chain is only as strong as its weakest link. These are prime examples of sales and marketing people forcing the relatively dumb consumer to focus on something that is largely irrelevant. It would be far better to upgrade low quality connectors and wires to the circuit boards first. In bad power areas, get a real line conditioner instead of an expensive power cable. Even in good power areas, a line conditioner can help as the power supplies in consumer equipment are often pathetic and under designed.
On the geek end of fixing, some people go as far as to replace the consumer grade IC's with professional ones. This comes from consumer, prosumer, and even professional grade equipment coming across as weak and not performing as advertised. While with video chips this isn't so practical, a lot of the audio crowd will replace the audio chips with something much better. Same goes for power supplies.
"I bought a super expensive speaker cable and my sound is clearly a little bit better using the scientific method to test it." No arguments there. The real problem was most likely the amplifier's output being low quality and having trouble with regular cables when it shouldn't have (probably capacitance). This is very common in consumer grade equipment. These principles hold true for just about any output (audio, video, RF).
Manufacturer of expensive cables often recommend running their cables the same maximum length as the cables from the discount place. A higher quality cable should be able to go further than a lower quality one.
Buying the same length speaker cables for "time alignment" when they are obvously not needed is moronic. Electric signals travel a high fraction the speed of light. That cannot be compared to the speed of sound. Time aligned cables (all the same length) really only apply to component type video signals that operate in the megahertz range or higher.
Using component video out instead of composite helps remove frequency interactions that can cause noise and dot crawl. Component video suffers less signal degradation over long cable runs.
Making a long video cable run is possible with YC cables made from good coax, but the luma and chroma signals will be skewed because the different frequencies travel at slightly different speeds. Professional video hardware usually has an adjustment to realign these.
Fixing a dirty ground (sometimes by adding a grounding cable) will do more than buying an over priced and over rated cable.
Some of the best speaker cables I've used have come from either generic lamp cord or a cut up extension cord. These are much cheaper than speaker cables and conduct just the same or better since the speaker cables often have less copper for the price. Get the soldering iron and tin the leads of the stranded cables. This lets the solder flow into the braid/twist and fills in all the dead space making contact with every little wire thus increasing overall connection surface area available to the connector. Reheating the solder is also an easy way to clean off the oxide.
Final: If you've read all this and still insist on blowing your money on something over price and over rated, don't come crying to me when you can't tell any difference.