Jeff,
that's a good calculation there. If the grinder is mounted to a bench, then a 1ft lever arm will probably give plenty of torque (total guess). For my flour mill, the flywheel is set up for a chain, so you can power with a stationary bike. I'm not sure I'll need to go that far for coffee: its a lot more friable than hard wheat.
I've also been wondering about roller mills, which are preferred at the industrial level (see mpechicago.com; they have a really good page of articles and whatnot:
http://www.mpechicago.com/coffee/...esource-Center.htm). Have a look at the articles. I'm not above buying a few test sieves (I already have the 40 and 60 mesh) and making some crude particle-size-distribution charts. If MPE is right, and roller mills grind TOO precisely (see their comments on "plurimodal" grinding, their term for mixing streams from two different roller grinders to make a blend), this could be kinda fun.
To make a roller mill, I've contacted crankandstein.com. They are curious about how many of us there are willing to make our own hoppers, &c, to make it worth doing a batch of rollers. If they decide not, it can't be that hard to make crappy ones on a lathe. Somehow, I think the Robur handmill will do better....
I think a micrometer head or other well-marked tool for adjusting the gap would be cool, although I'm not sure how to do it yet. The Zass design just uses normal, plastic bearings and a tensioning nut on the main shaft. Anyone know how to make this more precise than the GROF FEIN marked on the Zass bench/grain grinder?
best,
raj