dis-similar bicables? Why not?

I've been using NBS Statement speaker cables, but they are not bicables. (between a Pass amp and Revel Studios) Of course the speakers are setup to allow biwiring (and its recommended by Revel). I didn't want to spend another $3K to bicable, so decided to try using some Tara Labs Decades for the bass. I felt the bass is not as important, and removing the bass from the treble cabling might be beneficial. Results: I think it sounds better. I think the highs are a little more detailed. I'll listen for a wekk or two and then go back to listen again. Anyone out there try this?
I used different brand cables exclusively on my Magnepan MG3.3rs. I just made sure they were the same length. The Maggies always sounded best with better cable on the ribbon/midrange. The bass panel liked heavy guage cable..

Many Single Bi-wire cables use different gauges and composition for the highs and lows. Also, quite a few folks do this with Kimber 4(xx) and 8(xx).
I would think that with dissimilar cables, you might have a phase shift problem between the woofer & tweeter. But, I can't state that this is a fact. Better minds may be in order here! Happy Tunes
I wonder how much different geometries and dielectrics might affect the velocity of propagation at audio frequencies ? I know that Nordost speaks of a 94% velocity factor, but i don't know how "slow" or "fast" other cables conduct the signal. I know that this might be considered "unimportant" or "insubstantial" by some, but i would still be curious as to numbers. As previously mentioned, my main concern is a phase or time delay between the various frequency ranges.

Bottom line is, if it sounds good to you, that is all that matters. Sean
Sean, I think you're onto something there. When I used nordost red dawn on the newform ribbons it was too much. I think the speed of the ribbons worked against the speed of the cable resulting in extremely zippy notes with little body...I run the acoustic zen satori full range with mapleshade ribbon jumpers. This seems to give the right speed to both the ribbons and cone bass drivers. Interestingly, the virtual dynamics Audition power cord on the amp seems to add a "coherence and correctness to the sound. Is this a result of correct phasing? Or correct velocity? What artifacts have been removed? The point is that if it sounds great single wired, you can listen both ways to see what the cabeling is adding or subtratcing and as Sean says, go with what sounds best. It is possible that dis-similar cables on hybrid speakers can be promblematic.
The ulimate is on some YBA amplifiers. They have separate binding posts for biwire. On most amplifiers that do this, the signal is the same on both binding posts; it is just a convenience thing. On the YBA the signal for the mids / highs is more liquidy / lush, and the bass is "normal". If you don't like the extra lush highs, you just connect both highs and lows to the normal binding post, which is still outstanding.
There is no compelling evidence that two different cables will detract from the sound of your speakers. In fact, as many have testified, you can exploit the strengths of very different cables to achieve the best combination. It just makes sense that, as different as speaker cables can be, you would use the best, probably heaviest, cable for low end and best, probably smaller gauge, cable for highs. I even use a shotgun combination incorporating two different cables for the lows, albeit both are from same manufacturer, AQ. This seems to provide a lot of deep bass, with very good transient response. Double run of Slate gave very fast bass with no overhang, but not terribly deep. Double run of Midnight gave lower bass, but leading edge of transients was muffled and some lingering boom. Combined the two in true shotgun configuration, and voila! Deep bass with sharp attack.