Just WAGes of course from an engineer groupie here. The big boys in the big three design departments are probably using harmonic algorithms in their CAD models of the frame and whatnot. Those guys discovered they can reduce the vibrations of of the starter motor by .05% by placing it under the intake manifold and using breakaway proprietary sockets. (sigh).
Another WAG:
Pretend you are driving a rally race hard.
The engine within it's soft rubber motor mounts is torquing left/right/forward/back while the driver hammers and lets off constantly. The rear axle and differential is bouncing around in every direction as well over rocks and stumps. Both of these units are moving around at their maximum travel abilities. Randomly and fast. These are normally supposed to be pointed almost straight at each other. IIRC, it's supposed to be a 2-4 degree deflection angle for universal joints to work properly and for a long time I think. But instead, both their output-center-lines are mostly pointing everywhere.
A normal axle would just have two universal joints and a slider.
A center carrier adds a second slider.
But while I was thinking about it, you do seem to have a good question: By shortening the driveshaft into two with placing a fixed pivot in the center, aren't the drive angle stresses increased?
(unless that center carrier point floats around which currently doesn't happen
afaik)
Instead of a third bend 'around the corner', the carrier actually seems to be tightening the corner instead.