An obvious solution to this is doing exactly what you've elected: Develop the core, and the "add-ons" that are low hanging fruit (eg, those you have the hardware for), leave the kitchen-sink to plugin/mod makers.
Bear with me on the next part, it's a bit of a brain dump:
Objectively speaking, what stops you (or someone else in your situation) from using the community support you've received on your closed source app, and turning it into shareware, or worse yet, commercial? Those who have helped the overall end-product by donating money/hardware/code, end up with a piece of software they have no stake or control over.
Pragmatically speaking, the only real solution with a closed source "core" is to release these "sponsored" plugins under some form of license that is agreeable with those who sponsored. In the event that the "core" application does change license/cost, those who contributed (in some fashion) to the creation of the add-ons aren't left feeling "out in the cold".
Of course this is solely my opinion, and I wouldn't dare suggest that you (or Jirka, or anyone else) feel obliged to do it this way.
Frankly, I am more than happy to purchase the first "decent" Mac based system that comes down the pipe. I just don't see any available right now. And understandably so, the OSX space for this niche market just can't be a profitable platform for 99.9% of the companies out there. OEMs don't want to mess with Macs because of pricing and distribution issues, most users probably feel the same, and devs face some real challenges in terms of COTS hardware compatibility / community interest.
Which is how I end up feeling that the only way this is going to happen is through a community driven effort *even if it's towards a commercial application*.
So, after all this musing, meandering and rambling....
Here is my suggestion, based on all the excellent feedback I've gotten on my (flawed) idea:
Developers: Create bounties for your existing platforms.
NeonDev, you mention that you are interested in working on OBD-II and GPS, but you require the hardware to do so. Understood.
I will buy you an ElmScan/5 Bluetooth w/ USB/RS-232 converter, provided I can impose a few requests:
When you are able to achieve your goal of an OBD-II plugin/interface, the source for that plugin is released under GPLv3*
I also think it would be great if you created some sort of donation page so that people like myself can help contribute money towards other hardware/tools required for your ongoing development. Hopefully, work resulting from these community contributions would be similarly licensed.
It's likely not terribly portable due to it's heavy reliance on the Nokia tablets Hildon/GTK, but I imagine the core math and tiling code is sound.