i'd start at www.alumilite.com and look into water-clear casting plastics and associated dyes. That covers getting a transmissive plastic part. Now, masking... as i mentioned, half baked so we should probably rethink that... or at least test it. I was thinking if you could just get the tiniest bit of solvent under the mask, then it would pop off, despite being coated in paint. Obviously, the least possible amount of paint to do the job is ideal here. Well.. i'll keep pondering it.