>>3
I'm an engineer, not a theoretical scientist.
I think considering the potential temperature shift for a low power app like this we can probably treat the cabling as perfect, not like the silly non-ohmic conductor it really is. I don't even want to think about other effects like signal frequency component distribution across the cable diameter.
>>4
For a small MP3 player I don't think I'd want to use anything longer than about 5 metres, although if you're using good quality gear it might be able to hack a bit more. The longest I've tried successfully was about 8-10m, but that was fed into a stereo anyway, so the signal could be reamped.