If you want to know the true power of mylar, check these out:

http://www.diy-cockpits.org/coll/

http://mirrorsheeting.com/

Sweet. We used mylar (I assume) for parabolic carbon fiber concentrator mirrors for the solar car I worked on in undergrad. We never quite got that good of a finish. Putting the mylar on e.g. balloon rubber and anchoring that would be a fun experiment.

Another thought is I wonder if muscle wire could be used behind that, for a more compact setup than pneumatics.

The biggest problems I see with muscle wire instead of a cylinder and piston would be a) it would only pull wherever it was attached, instead of a smooth curve and b) it wouldn't be able to go convex. The cylinder(s) and piston(s)

can be mounted anywhere and connected to the pressure chamber via tubes. You could even build multiple cylinders on edge into the frame.

It would be convex in the off state. But in general, it would look "ribbed" probably.

Alas, I've already got to many projects I want to work on, and failing to focus on even one of them :P

I know what you mean. I'm part of the startup of a local makerspace/incubator, and all I'm probably going to be spending the better part of the next year designing and building tools. (...and tools to make tools.)

Even better: make the back mirror a sheet of mylar, with a cylinder and piston behind it. As a motor slides the piston back and forth, the changing air pressure will cause the mylar to go from concave to convex to concave.

Nice... My knowledge of optics is too fuzzy to be able to say how different the effect would be with a mirror, vs a simple reflective surface.