Dan, I was looking into this, they recently updated the $99 devkit to include not just the $495 assembler that it included before, but they now throw in the full $1395 c compiler as well, not the free demo one that's limited to compiling 4k of code max. I'm not sure if it comes with the floating point library which I'm pretty sure is necessary for your build and formerly was included with the $1395 compiler only, but I think it does come with that in the new bundle. If not, all we really need is fixed point anyway and not the full IEEE 754 fp standard their library implements, so if we had to we could build a simple fixed point library fairly easily.
Anyway, I am probably going to get the dev kit just for the heck of it because $99 is less than $1395. I agree $1395 was not a reasonable amount for hobbyists. I bet they sell more chips with this new devkit configuration.
So, if possible, it would be really nice if even if you don't do the whole firmware because of secrets if you guys could publish an address and i/o port map with the sensor mappings. Or maybe a barebones skeleton - like the code to run just raw sensor mode, which would be a pretty nice starting point. Something to think about. I'm totally happy at the moment using it in raw device mode right now with my 14 bit (just extrapolated from the center of gravity) controllers on the horizontal and vertical sliders. Since the device has to be run tethered to a USB host it's really not much different than what I'd do with firmware access. Raw mode makes it really easy to fool around with and do neat new things. But it would be also be cool to be able to in the future write directly to the flash and reconfigure it into a new device. This opens the possibility of more technically minded users to create totally new performance modes that just take over the whole system.
BTW, whether or not we have this, now that I understand that the AtoD convertors are 10 bits and there's possibly 10 bit pressure per sensor coming in though, it would be really helpful to have access to that for the pressure in particular since I run the aftertouch range up in the head room and with the current 7 bit info I have to lose resolution doing so.
Oh also, instead of toggling, could you please do on and off messages for the raw LED control mode? The problem is with the toggling since we can't query the device it's not possible to know what effect toggling has.
Thanks for letting us know about the chip, I think low level hacking could be a lot of fun.