Nope -- your right ... most all set it by adjusting alsamixer or pavucontrol (linux - Ubuntu) the Tx soundcard output level to what they need.
with that said .. it gets very tiresome ... when different bands have different drive requirements ... and even different amps -- hf,vhf/uhf. (UHFSDR)
The idea is to set the soundcard to a max level for the lowest gain band and have Quisk software lowering adjustments per band (maybe even mode too -- i run less power on FM than SSB on 2M)
Just now though of it -- even lowering the power (volume) for the Digtal modes might be a good idea for 100% duty cycle modes ... and FreeDV -- i always lower the Tx soundcard drive level there too ....
Did you invent fully different RF input and output stages for the UHFSDR?
The standard UHFSDR uses ADE-2ASK mixers with 1GHz max.
While the HMC 482 and HMC 544 are good for over 4GHz, the BGA612 maxes out at 2.8GHz.
Further, according to measurements made by me with a VNWA long time ago at the time of introduction of the UHF, the circuit board of the UHFSDR has shown rather reduced properties beyond about 600MHz as regards reflection and stability.
It therefore might be interesting what type in input and output stages you have devised.
Possibly the Lime SDR might be a better choice for your efforts.
Basically it states that the 3rd , 7th 11th harmonics are 180deg out of phase with fundemental ... and that 5th, 9th etc are in phase .... meaning that 3rd, 7th and 11th harmonic need I/Q reversed but 5th, 9th do not ...
think ill try this tomorrow with my UHFSDR .. by changing the controller divider to 1/5 of what it is now .. i should get a Lo that would normally be used for 10m .. but shows 2meters on the display and with a bandpass filter for 2m should get operation of 2m with I/Q in the correct phase.
I did this once before for 3rd harmonic -- not knowing about the 180deg phase change -- left me baffled for a bit as to why the I/Q needed reversed .... lol. It did work bty ... the If does not get changed in any way that i could detect.