NavList:

Peter you wrote:
"I'm not sure side-to-side error would be of much concern, though.  It would be something like side error in a sextant, I think---second order.  Would still need to keep it under reasonable control, since the turning may not be through exactly 180 degrees. "

I agree, but if the perpendicular "side" tilt is large, say, a degree or more, then it could have a significant impact since it's really unlikely that the mirror would be rotated exactly 180 degrees. 

All of these rotations, clever as Greg's trick is, are just ways of fixing a surface that isn't level. I feel like we need a pourable mirror surface. How about some of that mirror spray paint applied directly to a water surface? Has anyone played with that stuff? Or mirrored mylar floating on water? Or perhaps we could reduce the molecular weight of Mercury? Maybe a boiling hot bowl of zinc? Zinc is cheap, beautifully reflective, and it will melt over a natural gas frame. Oh but then there's metal fume fever... Maybe not! :)

Frank Reed
Conanicut Island USA