Mustafa
If you can address each point, which should be in the nanometers range, I would think you could do what you are trying to do, with a transparent polymer. You might want to look into UV lithography used for making ICs but that could also present a problem if the liquid medium is UV-attenuative, as it might very likely reduce the depth of polymerization needed for a large lens structure.
Are there any rapid prototype shops where you can try it in the macro scale? Even if you could make a single element it would be worthwhile, I think. What if you could have it reverse back to the liquid state, then reconfigure at another focal length. The OSA has a lot of papers on adaptive optics that might interest you. Microsintering of nanopowders is also very very intriguing. Imagine if you could have a monolayer of nanopowder with a below-threshold sintering energy present before taking it over the threshold with a projected image! I think this could be done with powdered glass using your idea.
Also, are current photochromic glasses polarized, or can be? That way you might modulate them with crossed beams. It certainly should be possible to project and hold an optical image on a sheet of photochromic glass, but I wonder if it would be blurry.
Bob