gsilver, that's absolutely right. Which is why it's necessary that longer runs require a lower gauge wire to compensate for signal attenuation. Back to basics.
BlackbeardBen, could it be possible that difference you heard was from the replacement of possibly internally corroded cables, or that replacing the cables disturbed contact corrosion that had developed over the 20 years? Removing such would make a drastic improvement on SQ. I think this is the case for experiencing improved sound with new cables, more than anything else.
I think internal corrosion could be a contributor, but I had fresh wire exposed on them (recently installed) so contact corrosion (at least at the connections) should not have been a problem. The wire had become quite stiff though.
I think the bigger issue may have been the length/gauge of the wire and coiling the excess (many loops, could have induced RF interference and/or enough back-EMF to notice). Remember, I was running 4 ohm nominal speakers that extend almost as low as 1 ohm at some frequencies, so the resistance of such a long and thin wire running such low impedance speaker at high power levels may have been a factor.