This is where I find the semi-conductive dielectric to be interesting. But you can't really have this discussion in public without being stoned.
I assume when he talks about timing errors he means group delay. One f is delayed (a different Vp) than another). This may be true but in the audio band 20-20,000 how much? Vp .= 1/(c sqrt(L C) L and C are essentially linear over that range. But even if Vp varies by 0.2, which I doubt that is 50 pS delta for a 10' cable. 20 MHz u wave realm impact? Yes. Audio? Imo no Even if it does happen you won't know it was 'wrong' because you have no reference, it will appear 'correct'. Even with reference imo you could not detect it. You can see from the equation if you increase L, C decreases, so they offset and the result varies little. In xx mile power transmission lines it matters. I have studied this subject. Assume a fault on a 50 mile line, if the Vp is 0.5 you can calculate the time it takes to dampen to a safe level to reclose a CB. It bounces back and forth 10x in 5 mS, based on line parameters you can determine losses/damping, each travel length. I'm not saying cables should not be carefully selected. But some of the things that are being said that should be considered in the selection process are not relevant.
People frequently mention things like skin effect that are not relevant in the range of audio frequencies. So many other factors may come into play in long transmission lines that have absolutely no meaning at short speaker cable runs.
