Optics isn't that hard. The microinch tolerances on surfaces are deceiving; you can easily verify and measure radii and surfaces to the required tolerances with homemade tools depending on standard manufacturing precision. When I made my telescope mirror, my Foucault tester depended on common 1/4" 20tpi all-thread rod and blind nuts, nesting sizes of hobby brass tubing, a razor blade made for a paint scraper, and model aircraft plywood. With it, I was able to figure the mirror to better than 1/5 wavelength of the yellow LED I used in it. It required care and patience of the same kind that leads from a straight print to a fine print (albeit applied over a longer time frame, given I had to wait 24 hours for the lap to press between polishing sessions).