This sounds a bit like part 1 of the 'uncommanded descent' issues being discussed in other threads. It might be that the firmware upgrade is coincidental - but ... A couple of questions and suggestions:
1) Have you noticed any difference in the sound of the drone? i.e. does it sound like it's rpm is higher??
1a) If you have a second set of prop's and can change the REAR set of 4 blades only, does that change the flying sound to a lower pitch [rpm]?
2) Have you noticed any difference in the stability of the drone in the hover? Does it hover steady, or wobble about a bit??
3) Fold out the MM's arms and prop's, and looking horizontally at the prop blades - do the front set have a different shape to the back set?? i.e. do the props on the back look a bit 'flattened'?
4) Do you store the MM folded up in a Fly More case?
5) If the answer to 4) is Yes - do you make sure that the prop blades on the rear arms are lying east/west across the body of the drone, or do they angle back toward the rear of the drone when it is packed in the foam cut-out?
There have been reports of the MM being stowed in the Fly More case with the [rear] blades that end up on the under-side, angled back toward the rear of the drone. If that happens, the drone will sandwich the blades under itself. A foam block on the lid of the case puts added pressure through the Mavic down onto the blades sitting on top of another foam block under the MM. It seems that if this stowage is not done carefully, the blades on the rear arms, can get 'flattened' and lose some of their pitch. That then means they have to over-rev to get the same lifting ability. It's possible that the over-rev can also make the blades stall and lose all lift, making the MM descend.
Ideally, when in the Fly More case, the rear prop-set should sit free in an open space in front of the foam block that supports the belly of the Mavic.
I'm wondering if the motor speed warning is the first stage of this problem that has been seen by a few Mavic Mini owners.