I wonder how much out of phase it takes to hear the difference. If it were only a small portion of the actual wavelength, then the number would get smaller, of course. When thinking of the system as a single sine wave, it is un-likely, but what about the fact that the system is an extremely layered complex sine wave? How would this minimally existent phase shift change the result?

You would also have a (very small) change in resistance, inductance and capacitance. How would these very small changes effect the flow of energy? Would the resistance of the longer wire change the parallel flow that much? Would the capacitance of the shorter wire change the reaction of the crossover? Who knows.

A gut feeling would be to say that none of it really matters at all and you could indeed have a 1 foot wire running in parallel to a 50 ft wire. But if some people hear a difference, and they might, then why not just say forget it and get speaker wires that are the same length. Its not that hard and it in fact would be insurance, as you have reduced the variation of more properties, even if that variation would be nearly impossible to quantify.

S.