Mhhh I did some extensive research on this bi-wiring thing 1 year ago.... and at that time I found an article with a PSPICE simulation which was pretty persuasive in explaining that what you'll get with bi-wiring is altering amplitude/phase of signal hitting the crossover filters, around the cross-over frequency. Unfortunately I cannot find the article now and I can't be bothered learning how to use search in the "web archive".
I've also did some experimenting with long runs of cable (15 feet as as to make differences easier to discern) on 2 different sets of speakers - Energy RC10 and Paradigm Studio 100. What I think I heard was supporting the article's finding.
To explain it simply:
-when you are single-wiring, the voltage at the entry point of the high-pass filter is the same as at the entry point for the low-pass one (as they are shunted together at the speaker terminals). The crossover will work as designed (and if competently designed will send signals of proper frequency, amplitude and phase to the tweeters/woofers).
-when you are bi-wiring, THE WIRES BECOME PART OF THE CROSSOVER FILTER. As now the point of same voltage is at the amplifier's output terminals. It would be equivalent to the situation where the amplifier is connected to the speakers terminals with an ideal 0" length cable, BUT you add 2 long runs of cable INSIDE the speaker, from the speaker terminals to each of the crossover filters input.
So next time a speaker manufacturer uses language like "DESIGNED for bi-wiring", ask them if the matching cables are included in the box. Since in a bi-wiring arrangement they are part of the crossover filter, they should be of an exactly specified impedance as to ensure that the crossover works as designed (the cable impedance would be part of the design).
Other speaker manufacturers are more prudent and use milder language like "allows bi-wiring" or "compatible with bi-wiring" or "2 sets of terminals for bi-wiring applications". That's OK with me.
Nap.