I know what you mean - the multiple alternatives are great for
innovation but we do suffer fragmentation compared to other open source
autopilots.
I agree, but it is normal in my opinion. Today making HW is relative cheap and
chips are more "stand-alone" than in the past (you really do not need to be an analog/digital expert).
So, if you plan to modify a board for fitting your HW requirements, it will take little time.
- Same processor as the tiny. Known good and tested firmware, the layout
will be simpler because this board is an evolution of the Tiny.
Here I agree with Antoine, today the "world" is moving (or it has been moved) to Cortex.
And it seems it will be the new "standard" for the next following years. New designs,
in my opinion, should move to this architecture.
- The IMU is onboard. Not on a daughter board, no extra wiring.
It depends on your vehicle of course, sometimes it is useful to have the sensors in another board,
specially the magnetometer and pressure sensors.
- The performance of this board is not as good as the Lisa.
But why do you need performance? If you are doing crazy control/data
fusion work that absolutely has to happen on your mission-critical
flight processor then you are probably in a minority. Most stuff can
happen on a separate off-board processor, like a gumstix.
It is a fact that you only uses the power when it is available. Of course
if you only need the basics, you can even use an 8bit microcontroller.
Today CortexM4 is out, with floating point unit and DSP instruction set, maybe if you need a heavier
GNC you can implement it there without needing an ARM9 + DSP extra :P. In my opinion this is
a great breakthrough.
Moreover, new micros come with new peripherals, you can interface an SD card
with a high transfer rate by HW (I was waiting all my life for cheap ARM7 with this feature xD) , they come with better ADCs, DACs, etc.
And of course, without changing the IC package, conserving your layout and board dimensions. But like always, It depends on your requirements.
Cheers,