I'm no engineer, clearly. But I also wondered if the IBIS cradle/system made the camera more prone to shutter shock. If so, perhaps some minor adjustment could reduce it, but that would mean sending in an otherwise perfectly fine body. (OR, what if electric current is involved in keeping it "parked" when camera is on but IBIS is off...Could firmware address that?)
I think the testing done by canonnews is very convincing, comparing various EF-M lenses with mechanical and fully electronic shutter on the M6 II. To me, it makes it clear that at least for the EF-M lenses, IS 'OFF' is not mechanically locked. It is simply a setting in which the forces acting on the floating lens elements attempt to keep them fixed relative to the lens body, rather than the world. Which is not that surprising, as I believe the common wisdom is that EF lenses behave exactly the same way.
But lens IS is different from IBIS, so it could have been that Canon actually has some form of lock in the IBIS unit.
I haven't thought of a good experiment to show that movement in the sensor rather than the system as a whole is the source of blur. But it does indeed sound reasonable (which is not enough to treat it as fact).
Measuring contrast rather than looking at images as done by canonnews seems much more robust methodology, as it would also allow detecting the shutter speed at which the blur is most extreme. If the motors are just applying enough torque to the body for it to move enough to degrade the quality, I would expect the most affected shutter speed to change on a lens per lens basis. After all, as the mass of the system changes, so should the time it takes for the shutter to
So far, it sounds like everybody who did report in the 'issue' has the same experience with around 1/100 th of second though, right?
If the sensor moves despite the IBIS being off, could you attached something to the flash shoe or tripod mount so that it is in the frame, but static relative to the body? If so that subject should be sharp regardless of shutter speed and motion of the camera (as long as it is attached rigidly enough). If it doesn't, it either was affected by moving the camera (not stable enough) or the sensor inside the camera was moving, blurring the static subject in the process.