Moving reflectors is a very interesting idea. I wonder if you could use it to increase the amount of light falling on the panel, kinda like the way they made a "morphing parabolic concentrator" for the Solar One plant in southern CA.
(Here's an extremely dry engineering description of the mechanical details for the terminally geeky. For less-than-terminal nerds, it's worth doing some googling to find good pictures of it: I've driven past it a few times, and will always regret that I didn't take the time to do a good photo shoot of it. It's really quite beautiful if you're the sort who finds suspension bridges and GT race car designs "sexy").
You might also be able to significantly improve efficiency by using reflectors with coatings that absorbed wavelengths that are turned into waste heat, instead of electricity, by the panel.
But, in many cases, you won't really save anything mechanically: where I live, high winds would require the reflector movers to be almost as robust as ones to move panels. Maybe even more so, because the reflectors could wind up being pointed in directions that have higher wind loads than simple east-to-west tilting of fairly-horizontal panels.
But, on the other other hand :), you could make the reflectors somewhat porous, and trade off wind loading against some lost reflector efficiency. Hmmmm...
Maybe we should look into filing an R&D grant application ;D
Ran