RT3S / MD-UV380 too little frequency swing on 2m FM

[Derecho]

This issue I was initially made aware of by another ham, rather than discovering it myself. Since obtaining the TYT recently I've been testing its features out and also tried to get onto the local, 2m, FM repeater. They reported not getting anything coherent at all. I quickly grabbed my Baofeng UV-5R which I had nearby from earlier testing, and whilst I still was not amazingly clear, they now reported they could understand me and we had an actual conversation. This ham used some undefined equipment to measure my signal on the repeater's input frequency as he received it, and I tested the TYT, Baofeng and a QYT in my car as well this way. He reported the QYT to have the best-looking, strongest and clearest signal, the Baofeng to be a little low/noisy but otherwise sufficient and workable, and the TYT to have much too little swing.

After this on another day, I went to do some tests myself. I set up an rtl-sdr in my attic, and used the TYT and QYT from the ground floor alternatively, looking at my signal with gqrx over rtl_tcp. I've done various tests and tried some factors (with/without CTCSS, narrow/wide bandwith of RX filter, programmed mic gain), but it all comes down to the following conclusion:

On 2m, the TYT produces visibly less swing than the Baofeng in all the tests:

(On this screenshot the mic gain is set to: FW setting 5, qdmr setting 8. But also with FW 4 and qdmr 6 it seemed the same really.)
In addition, the TYT's demodulated audio signal (that helpfully comes with a delay due to my setup) is subjectively perceived at a much lower volume than the Baofeng's.

Curiously, this is not the case on 70cm:

The swing looks very similar and the signal perhaps slightly stronger.

I can also report that the stock firmware (RT3S 19.26) does not suffer from this issue. Like with OpenRTX on 70cm, the signal appears stronger, in this case more so even: