Most of the time my mantra is "keep it as simple as possible" also for my strategy in making things. The earlier shown second version of the rotary drive was build in three parts some of them from brass because it's easier to handle on the mill.
The drive works well under lab conditions but not if some cross forces comes up. If I increase the clamping to handle the forces this happens
Major problem was the lack of a baseplate.
To build the complete lower part from a massive block would be too expensive.
Thereby I've decided to build it from stainless parts and weld them together.
Works now perfect with all forces up to a stalled motor.