I saw that too in RC Groups, I just don't buy it. Seems VERY far fetched. As protected as these craft are in the Styrofoam boxes, it's extremely unlikely a twisting force strong enough could occur to twist the gimbal with heating and cooling. That's almost laughable IMO. And physically twisting it fix it? That's a pretty bizarre prescription to resolve, for such a delicate device.
Here's another reason I don't believe it. DJI recently repaired my
P4P for a bad gimbal in Cerritos, CA. When they replaced the gimbal/camera assembly they didn't bother to check it being straight before shipping it to me, just like the factory! When I got it back the camera was crooked! I live only 30mi from DJI Cerritos, UPS ground is overnight shipping. There's no way the "heat up >cool down story" could apply to my experience. That's why I don't believe that RC Groups explanation, at all. That's BS IMO. This is a process problem, a training problem, a quality consistency problem.
I fixed my crooked camera with a
dual axis gimbal calibration (that process isn't covered in the manual). I didn't have to man-handle the gimbal and twist into place. It's been fine ever since.
Don't get me wrong, I love DJI products, always have. It's just frustrating they don't see the virtue in a thorough outgoing inspection process to mitigate unnecessary warranty returns.