I accidentally bonded the achromatic camera 1/2 doublet in backwards. Camera 1 is a biconvex lens with a slightly different radius on side 1 than side 2. Yes, I _thought_ that I took all precautions: I compared the radius of surfaces by eye, and I also saw interference fringes from the matched ROCs of both elements. In retrospect, I should have marked both lenses so that during the application of the PDMS (aka Sylgard 184) I would not be able to make this mistake.
Never-the-less both camera 1/2 were bonded incorrectly. The project cannot continue in a meaningful way until this problem is solved. I have now decided on taking two simultaneous forward paths and will use the path that will produce results faster. The most promising path, at this point, is to attempt to debond the doublet. Here I'll use toluene which turns out to stretch the sylgard-184 and cause it to delaminate.
Thus, the next week and a half are now dedicated to fixing this problem. First light is six weeks away.