Just like everyone else, I have my theories. For once, I believe that temperature is a factor but not the heat of the motors - quite the opposite in fact, as low temps do have a worse effect on polymers in general (of course considering both to a certain point). Plastic becomes more "glassy", more brittle and that could lead to the start of fractures in certain structures and the P3 shell is a monocoque structure.
Second, I am not sure the active braking has any more bad effect than any other sudden change in torque, at lest not in a significant degree. The motors act in controlled and concerted manner due to the individual ESC and that does mean they change torque and there are severe twisting going on on the AC during certain (and different) maneuvers. But not to the point of causing enough stress to crack, it should withstand that and much more as it does on seemingly most P3s. Besides, they motors are projected and built on a quad to adjust and compensate to drive the AC in ways that no single motor gets overloaded or overburdened in comparison to others for too long. And if that occurs then it might be caused by some imbalance in the motor/prop. Yes, prop balancing is that important, if you come from RC you know what I mean.
In aviation it's usually a sum of two or more commonly more causes that leads to failure, like cracks. Actually in engineering in general, no single factor can cause or explain cracks in any dynamic structure. It might be the case here.
Considering most P3s (either from newbies or experienced, bold or conservative pilots) perform similar moves, but in different conditions - and not all of them crack - IMHO these problems are caused by a wrong or bad mix in the plastic used to manufacture the shells in conjunction with low temperatures. Of course from different batches, I'd guess not even DJI or its supplier may be able to tell exactly because they're seemingly overwhelmed with the success of P3s and even though DJI stuff in general seems to the manufactured to a high standard, QC issues can be a factor when there are big outputs involved.
Now, I am aware that some crack even in the summer or warm climates (like where I live), and some don't. The apparent randomness of the cracks leads me to believe that the root of the problem is indeed in the plastic. Not the design (otherwise a vast majority would crack and doesn't seem to be the case), not the brake (some set at 100 and still don't crack, while others set at 60 or even lower do crack), and certainly not the motors being new or old, as both crack - though no one knows at what rate on each.