(Posted this as a seperate message so not to mix multiple subjects)
As you mention "microcontrollers in the signal-chain of a transceiver", I am currently looking into OpenRTX.
It is really a very nice example of exactly what you mention and something that has become possible to last 1 to 2 years.
With these radios that support opensource firmware, It really has allowed amateurs a look of what is inside of the firmware of a "commercial-grade" handheld radio.
Two weeks ago, I helped out in an infobooth on Amateurradio at a makerfaire here in Belgium. Things like OpenRTX allow to explain to IT-people (who normally only work on computers) how "embedded software" works, how software that runs in devices we use everyday operates.
In that sense, FOSS is as much an educational tool as it is "just a piece of code that does something".
You introduce a really interesting idea in mentioning the OpenRTX project, that's not something I've come across.
I would like to point out however that as far as I know, the only open source hardware that this runs on is the M17 module built by the M17 project. It's my understanding that the other hardware has been hacked to be able to install open source firmware onto it, but I suspect strongly that this would void your warranty.
Consider for example the situation where a firmware update failed. Unless the firmware on the radio and the firmware you were updating are both supplied by the manufacturer, they're unlikely to assist you and I doubt that any legal requirements exist to make them.
I'm making this point because that distinction might be important to an end user buying a new radio with the potential of bricking it without manufacturer support.
I run OpenRTX on a Retevis RT3s, which can be done without any hardware modification.
(I do not know if original firmware is available somewhere -I have not checked-. If that is the case, it should be possible to reflash the stock firmware on the device).
Anycase, I must say that M17 does not run correctly on that radio.
There seems to be an issue that the first 300 ms or so of the transmission is not correctly modulated (something related to the FM modulator) and also the end of the transmission is broken of halfway the end-of-transmission frame.
I am currenly at the stage of trying to understand how OpenRTX really works, and my first idea is to implement POCSAG-paging into it. (As I have the source-code for that here anyway) and I also have some ideas for APRS to want to delve in.
(OK, that is, if I have some time left next to all the other stuff I am working on :-) ).