Not entirely sure to what extent you agree with me!
I agree entirely that science does provide mechanisms whereby cables can affect the sound, just not the magic ones beloved of audiophiles, that are invariably dependent on some unknown aspect of the cable. As for my foolishly adopted stance, it is one that I feel that one must adopt by logical necessity when faced with unfalsifiable claims with no supporting evidence, lest I seriously entertain the possibility that I am followed by invisible, incorporeal gnomes. See http://en.wikipedia.org/wiki/Russell's_teapot.
Or is your criticism of pt.3 (and indeed 1) more a squabble over semantics? Technically, some audiophile ideas about cables make sense (sufficient gauge, that sort of thing), but I think you know to what I refer.
We agree in that it is not magic. We also agree that explaining any change as being caused by magic is useless.
It is very important to seperate the claims made of the cables themselves, and what happens when they are used in situ. I can make two different IC types which will measure identically on the bench to 7 or 8 digits precision, yet perform very differently in the final application. If I did not know why, then I would be left with magic as the only plausible explanation.
Tossing out an individual's assertion of a change in sound because they cannot present a reasonably coherent explanation is what I call foolish, or better yet, just unwise. It doesn't mean they are correct in their assertion, just that it is not wise to toss it because the person does not have any engineering understandings. I extend that lack of expertise to the vendors as well..look at the shunyata test...well done, may indeed correlate to a change, but has an explanation with no meaning in reality. I would not toss the possibility of a line cord effecting a difference just because I know their explanation is bogus.. They claim cord characteristic impedance is the thing, but they missed the boat.