I've been out of town and I'm coming in a little late here, but there are few things in audio that I understand LESS than the supposed advantages of bi-wiring. Which, if I understand correctly, consists of running two parallel sets of wire from an amp channel to a two-or-more-way speaker. At the speaker end, you are still sending the signal into a passive crossover of some description, right? So nothing of substance changed there by bi-wiring. At the amp end, the wires are electrically connected to each other, right? Then logically, the only even theoretically possible advantage is lower resistance, yet as others have said, a single 10- or even 12-gauge stranded copper wire pair with decent insulation will measure negligible resistance to speaker loads over any reasonable run length. On my upper-level-consumer-grade ohmeter, I measure zero resistance on my six-foot run of 10-gauge stranded copper.

If "collapsing magnetic fields from bass signals ... induce distortion in the upper frequencies", etc. etc., this is going to happen anyway as bi-wiring is neither electrically isolating the signals nor physically moving the upper frequency speaker inputs further from the bass.

As others have said, perhaps those who have noticed improvement have been using crummy wire to start with (I can believe that bi-wiring lamp zip cord, for example, could measurably lower resistance compared with a single pair). VERY long wire runs might also be occasion for bi-wiring. Another possible reason for perceived improvement is the simple act of changing the connection. I believe that most of our friction-based audio connections degrade electrically due to minute amounts of surface oxidation. Every couple of months, I simply pull out and then reinsert each of my connectors, and I would swear that I SOMETIMES notice a slightly cleaner signal.

On the other hand, I believe that depending on the speakers and especially the design of their crossover, one CAN achieve audible improvement by bi-amping, whereby the bass and treble signals are more or less de-coupled at the amp end. (For those with the 5, 6 and 7-speaker systems, I would say that audible results would only be achieved on the "main" speakers, and only then if they are set to "large" and are in themselves of high quality). Of course, unless the amps are truly identical, bi-amping does introduce a complex new set of paramters in terms of volume and frequency matching. I bi-amped for some years (roughly 1973-1988) on my two-channel music system, and all of my audio friends concurred that there was audible improvement in clarity and attack or transient response. I stopped because with my next set of speakers, improved and much more complex crossover design both reduced the potential improvement, and made it very difficult to accurately match gain settings on the two amps.