From what I have read, one issue with BLDC motors at low RPM involves "cogging", so it seems a GR would still be very much on the plate. What I am really curious about, though, is this:
(and this may not be the proper forum for this discussion, so I mention it only in passing!)
If I were to embark on a project such as developing a gear reduction such as John's or Robert's, I would first model it and then make a 3D print of my design for proving before committing the time and effort in doing the machine work. But then if I were to use some high strength nylon or carbon filled filament to produce my design, maybe that alone would be adequate, or at least adequate to the degree that I could use it to aid in machining the final item.
That workflow is what I would use if I were to embark on the ELS path. I think some of the aluminum parts could be 3D printed and then perhaps replaced at some point in the future, once everything is up and running.
