Yes, the transducer idea is very simple to implement... I was not aware it is being done commercially. As far as the higher order harmonics are concerned, only SOME will be blocked, right? If I remember my physics lectures correctly, the one after the fundamental has a node at half the string length, and the one after that has it at one-thirds, and so on. So the ones with the node over the driver would probably be in dog-whistle territory. I mean, except one in the lower order. So perhaps the second order harmonic will be dead, and all other after the seventh... not a big deal, the problem is aggravated with increasing string length, and I doubt anyone would want a fifth-order harmonic on the 22nd fret. But yes, it does make it unpredictable. That's where your project is promising- you could perhaps sense it, and have an alternate driver purely for this situation.
The floting bridge has other issues- for instance, I'd be completely blocking the driver when I pitch bend with the tremolo (damping). Also, it will probably make the other strings move- forced vibration. But I'm sure that happens with the magnetic driver as well.
My aversion is not due to complexity, but due to the output I desire- if there's high crosstalk, I might as well go for a blade pickup. But at the same time, I need the ability to bend strings without issues... plus I'm not going for the kind of thing you are- I like it barebones. So a traditional pickup makes sense for me. But then again, I have never used a hex pickup, so it is all speculation....
If I were to do it, I'd probably have some uC in the guitar to control it and mix all six into a single output for use with a regular amp. I'd also have a few switches for the extra "settings" and eq. or whatever. When there's a new one required, simply connect it to a computer and send it a new config-file. But hey, maybe that's just me
Very, very ambitious... keep us posted!