If "skin effect" has no effect at audio frequencies, try building some IDENTICAL geometry interconnects using identical materials BUT with the only difference being wire gauge. For the sake of comparison, you would have to use solid wire since stranded wire does not come in "micro" gauges.
For specifics, i'm talking about using some teflon jacketed solid wire of various gauges arranged in a twisted pair configuration. Start off small ( for instance, 26 gauge ) and then make a pair that is several gauges heavier ( 20 gauge ) and one other pair that is several gauges heavier than that ( 16 gauge ).
Given the fact that the materials are the same or very similar, the geometries are near identical, etc.. the electrical parameters should be "ballpark". Not exactly the same, but ballpark. The one that i know that you will mention is probably series resistance, but that should be less than 2 ohms total so long as the cable is not more than a few feet. Since the load that the source will see is in effect multiple thousands of ohms, this should have little to no bearing on the results.
Now hook one set up at a time and listen to them. Then swap to the next heaviest, and repeat the swap for the last set. You should instantly notice a difference in tonal balance, high frequency extension, the clarity and quality of high frequencies, etc... between these three cables. Overall, there should be a MARKED difference in treble performance.
If anyone HAS done "testing" like this, PLEASE post your results. Then try to explain why you think you hear the differences that you do. Keep in mind that "current limiting" has NOTHING to do with an interconnect, as we are talking the lowest levels possible. Of course, EE's and anybody else with a "theory" is welcome to join in. Just so long as you know and understand that it is a "theory" that we are discussing. Sean
>