I'm certainly no expert, but isn't what I described considered "active bi-amping?" i.e., using the active, electronic crossover in my AVR to split the signal to the woofers and tweeter. I would be bypassing the speaker's internal passive crossover completely.
http://sound.westhost.com/bi-amp.htm
What you are describing is indeed active bi-amping.
Here are my concerns for you if you do attempt it.
The first, most obvious issue is that, assuming the active crossover and all its utilities (noise generators etc.) do in fact function properly, you have no way of knowing, without either the original speaker designer or the original speaker design specification and validation notes available to you, what the relative levels, crossover frequencies, and equalization of the existing passive crossover have been tuned to.
Presumably the passive crossover has been optimized with at least some minimal design rigor and testing in a controlled laboratory environment. You will not be able to attain the performance of even the passive crossover without similar expertise, effort, and equipment unless the speakers are incredibly cheap and have only blocking capacitors on the tweeters and maybe iron core inductors on the woofers. If any design rigor went into those crossovers they probably outperform anything you can cook up through trial-and-error with the active bi-amp feature of your receiver.
You do have one factor working in your favor however. In-wall woofers are infinite baffle design i.e. frequency-independent half-space loading on the drivers (the wall divides the space in two) and do not experience loss of bass when the wavelength exceeds the size of the baffle and starts radiating omnidirectionally (approximately 1KHz in typical free-standing speakers) so they do not employ baffle step compensation on the woofers at the transition frequency.
http://sound.westhost.com/bafflestep.htm
So as long as you mount the speakers in a wall as intended you will not have to include that particular (baffle step) compensation in your active crossover (not that the receiver likely gives you that capability anyway...).
The point is, this active crossover is just a marketing gimmick and not really intended to be used as an active crossover. They basically expect you to use the bi-amp capable split crossover terminals of standard towers anyway just like you would for passive bi-amping, unless my guess is totally off the mark. I am not expecting you to be able to dial in baffle step compensation or tweeter compensation or anything like that, just a simple split into two bands with a tunable Linkwitz-Riley crossover frequency and maybe if you are lucky you can select the order (slope) of the crossover too but probably not independently for woofer and tweter.
The second issue is that active bi-amping often employs amplifiers with power sourcing capability matched to the particular drivers' power handling capability in order to ensure that no driver is ridiculously overpowered and put at risk from a grossly over-capable (and thus wasted) power amp. Using a separate amplifier channel of your receiver to power a single tweeter will definitely put that tweeter at risk with approximately 10 to 20 times its power handling capacity and no DC/LF blocking capacitor in the circuit to protect it if for some reason your receiver hiccups and outputs low frequencies to the tweeter. You could protect the tweeter with a blocking capacitor of your own but that sort of defeats the whole purpose of active bi-amping anyway -- to eliminate passive crossover completely.
The third issue is that if your speakers are indeed a two-way WTW LCR configuration, the total power handling capability of the speaker system is likely not so extreme plus the LFE/bass extension is likely already split by active crossover and bass management to a powered subwoofer anyway.
Splitting the tweeter/highs off that amp does not accomplish much given that each amp channel is probably quite capable of driving the entire speaker system, particularly if you already have active subwoofer. It is almost as if your system is already 3-way and your in-wall speakers are the mid/tweeter with your subwoofer taking the role of woofer.
Active bi-amp (particularly done ad-hoc like you propose) is more suitable to live sound in such cases where the total power required from the amplifier reaches such magnitude that design issues such as transistor/tube power dissipation and voltage peaks begin adding tremendous cost to the amplifier implementation for little incremental gain in sound level.
In such live sound applications, either the setup is a tuned preconfiguration, or the person setting it up is a sound engineer with the appropriate tools and expertise to get it right, or the band just does not have the resources to do it right and settles for whatever performance they can get at a bargain basement budget with Uncle Charlie taking on the role of acoustic engineer. If you want your home theater to sound like a band playing a dive, novice active bi-amping would be a good way to get there.
Because of all these factors, you will not really gain anything at all from bi-amping, with the following two (basically minuscule impact) exceptions:
1) any intermodulation distortion in the amplifier between low/mid and high frequencies will be alleviated by the active crossover separating some frequencies that could intermodulate (but the contaminating sum and difference frequencies that IM generates are probably not audible anyway unless the amp design is sketchy)
2) wire losses will be reduced (not an issue unless your wire is significantly longer than 50' and 12 ga or smaller and/or perhaps grossly substandard lossy such as Monoprice, Best Buy or Fry's as measurements by amirm have revealed).
The final consideration I can think of is that even though your receiver has an active crossover like my TX-NR929 does, it might not have full functionality in the firmware and/or hardware and you might not even realize that it is not working properly until it is too late and you have already blown a tweeter. In my case, the pink noise generator spectrum is supposed to split between the high pass and low pass channels but it gets confused and starts filtering neither/both with the same filter during its 'calibration'.
Fortunately I discovered this when testing with full range passive crossover speakers connected to both high and low frequency amp channels rather than after bypassing the passive crossovers, or I could have destroyed one or more tweeters with the malfunctioning receiver, or at least wasted all that effort trying to get it working just to discover that the manufacturer apparently never actually expected anyone to use that feature anyway and shipped it functionally broken.
An alternative explanation could be that the firmware in my receiver got confused and needed a hard reset issued from the front panel. I know that the tone controls were only functional on the main l/r channels at the time (they recovered after a reset) but honestly, if a firmware glitch can destroy my tweeters I am not interested, thank you!
In summary I would advise you to give up on active bi-amping unless you are a DIY speaker designer equipped with the necessary education and tools to make it happen properly (and your receiver IME does not qualify as the necessary active crossover if it in any way misbehaves like mine does). This is not something you want to jump into blindly. Speaker design is one of the most 'black art' aspects of audio as well as one of the weakest links in the audio signal path. Probably not a good idea to mess with it unless you know exactly what you are doing.
This assessment comes from a person with a BSEE and some long-standing but minimal impact DIY audio/video. I have just enough information and experience to be dangerous, or alternatively, just enough to realize that it is probably a bad idea to proceed without any formal training or at least deep prior involvement in speaker design.
If you really want to try out the active bi-amping function of your receiver, I suggest that you first try it out the way I did, with full range speaker systems on each of the frequency-limited amp outputs. If your curiosity is satisfied after doing so, maybe you will avoid any unpleasant misfortunes and/or wasted effort with your out-of-production speakers.
