
The degree of precision required for the encoder (.010 max runout) and lack of a suitable disk means that we won't see this design anytime soon.
Sorry guys.
I wonder if the principle can be applied to a spoke-counting design? I'll have to look into that...