I think you're seriously abusing the word modulate here. You can't "modulate" a signal with a impedance, even if that impedance changes with frequency.

Let's think of this a little differently. We have an amp, with two parallel sets of wires. One set goes to a crossover, then to a tweeter. The other set goes to a crossover, and then the bass. The wire attached to the crossover of the tweeter presents a high impedance to low frequencies, and a low impedance to the high frequencies. The wire attached to the crossover of the bass presents a high impedance to high frequencies, and a low impedance to the low frequencies.

At the tweeter crossover we will see almost no voltage drop for the low frequency signals, as almost no current is flowing at low frequencies! We will see voltage drop at the high frequency signals, as current is flowing at the higher frequencies. The opposite will happen to the bass, with no drop in voltage for high frequencies, and voltage drop does occur at low frequencies.

So what is the effect? Why might you hear a difference? A tweeter is much more efficient. Some low mass silk or titanium dome moving small distances, is much more efficient that a large mass trying to move a quarter inch. So the current in the bass wires will be larger than that in the treble for the same sound pressure level.

Now for the real question, does seperating the wires create a lower impedance for the mid's and high's resulting in a clearer mid/high. The concept of super-position is important here. It says that the response of the system is the same as the response of the sum of the components of the response. This is true, as there are no semi-conductors in the wire. There is no modulation, no mixing of the frequencies to create other signals in the frequency response range of the speaker. It is the concept of superposition, and the fact that the frequencies won't magically mix that allows us to have channel 3 and channel 4 on the same coax, neither channel interfering with the other. The current of the channel 4 signal doesn't modulate the channel 3 signal!

Super-position says that no effect will be seen. This of course assumes that you are comparing identical wire gauges. If you have speaker wires so that the wire gauge of the bi-wiring pair is the same as a standard single pair, you will not hear a difference. I contend that people hear a difference because they double up on the speaker wire when bi-wiring.

I've tried connecting two sets of wires from an amp to a speaker. First bi wiring to the designated locations in the crossover, and then just essentially decreasing the AWG of the wire by connecting both to the single input. I've never heard any system where the individual listening, or I, could hear a difference. (It's important to do this having someone else handle the wiring, and the person listening unable to see which way the wiring is done. Otherwise, you're most likely to prove what you believe, rather than the truth.)

So my position is that bi-wiring is equivalent to running bigger wire. If you find it more convenient to run multiple runs, and bi-wire, then go ahead. I'd even agree that it is reasonable as it allows for bi-amping later as an easy upgrade. Bi-amping does offer real benefits!