WIP: Omnibus MC only

[dagar]

Needed to unblock #10355, but this is also a good excuse to fix these bogus dependencies.

Navigator is currently dependant on the VTOL parameters VT_B_DEC_MSS and VT_B_REV_DEL. Let's either handle this with the VTOL controller publishing a status message, or falling back to the old param API.

I'd vote for an orb based solution if @RomanBapst could assist. I'll be mostly unavailable until next week.

Similar to the way the FW position controller publishes position_controller_status for the L1 distance that's used by navigator as the acceptance radius, we could have the vtol controller publish an appropriate acceptance distance for back transition (how VT_B_DEC_MSS and VT_B_REV_DEL are used). https://github.com/PX4/Firmware/blob/master/src/modules/navigator/mission_block.cpp#L291-L305

Going forward I think we should strictly enforce parameter module boundaries with only a few global exceptions (SYS and CAL?).

[dagar]

Crazyflie and the Intel Aero can be fixed as well. https://github.com/PX4/Firmware/blob/master/cmake/configs/nuttx_crazyflie_default.cmake#L65
https://github.com/PX4/Firmware/blob/master/cmake/configs/nuttx_aerofc-v1_default.cmake#L59

[bkueng]

I went for the simple solution (C-API) in #10738 to unblock things.