You would essentially need some kind of hybrid capacitive controller, and a multi-layered PCB so convoluted and expensive that it would defeat the purpose.
I'm sure it wouldn't be viable on a small scale but I guess what myself and one of my electrical engineers are scratching our head at is the notion of impossibility. We are of the thinking that the controller should be where the real focus should lie if someone were to do this on any scale (I'm not going to actually build a controller, but we're certainly invested in the thought of this now).
I guess I'm just confused because it looks pretty simple to us and we're in the business of building impossible (or as close to it as one can get). I mean... We've already got two drawings of a topre type switch in an mx-style housing after finding the novatouch variant and technical drawings (pretty smart design from a manufacturing perspective... Not sure about "feel").
Well if you're using an the word
impossible in the strictly lexical sense, then no, it's probably not impossible. I mean, consumer-grade 14nm semiconductors are possible, so let's change our verbiage here and say it's
implausible. I'll do my best to explain why.
A keyboard PCB for discrete switches is not an inherently complicated design. This is evidenced by the existence of hand-wired keyboards. Most of the PCB is empty space, with through-holes for one or more switch types, and usually through holes and/or SMD solder pads for diodes. (Let's leave out LED backlighting for now, because it will just unnecessarily complicate things, and Topre couldn't support it anyway because of the solid rubber sheet over top of the PCB.) The controller has a simple job: listen for closed circuits and fire the corresponding scancodes. Capacitive PCBs are much more complicated. They have capacitive pads instead of through-holes, and they don't have diodes since capacitance is unidirectional. The lack of holes and solder pads makes for lots of good breathing room for nice, big, round capacitive pads. That's what you want.
Capacitive PCB pads are funny things. Their shape and size have a significant impact on their performance. Once you start drilling arbitrary holes for MX switches, you affect the capacitance in unpredictable ways. Even assuming you could "tune" the pads to compensate for the extra holes, you still have the extra leads to deal with, since you can't double up the capacitive and digital matrices on the same circuit.
If you are talking about developing a discrete Topre-style, rubberdome-over-capacitive PCB switch, good luck. If you thought fine-tuning a production-grade capacitive PCB was hard, try compensating for amateur PCB soldering jobs or, God forbid, a hand-wired board. And what controller are you going to wire it up to? There are currently 0 (zero) hobbyist Topre-compatible controllers.
Assuming you manage to get all this working, who is going to buy it? The issue of keycap incompatibility has been addressed by the Novatouch, followed by a number of up-and-coming aftermarket slider initiatives. So really the only conceivable benefit to a fully-custom Topre board, is layout variation. And there doesn't seem to be much demand for that among Topre enthusiasts.
These are just the first things that pop into my head that make the concept of an MX/Topre hybrid seem highly implausible. I'm sure there are smart folks in the forum who can expand on these points, but if the idea were even remotely marketable I think it would already exist by now.
