Oh dear... lapping. How accurate do you want to go?
I've spent over half my life lapping and polishing to optical tolerances where you're talking flatness tolerances over (say) a 12" diameter surface of a twentieth of a wavelength of light. For you sheddies out there using micrometers, this roughly equates to 32 nanometres (at 633nm wavelength) over 12" diameter... that's peak to valley form error by the way, or departure from perfection.
You can't really measure better than this cos light diffracts after this and no longer behaves properly lol, you use test plates (look up Newton's Rings) or Interferometry (Phase Shifting Interferomery is better), but the old techniques used since Newton's time are still just as relevant.
Read F. Twyman's "Prism and Lens Making"- all the other books just copy him really.
(even he don't say really how to do it, but it's as close as you'll get, us optical folk keep the techniques jelously guarded hehe).
Bear in mind at this level of operation glass can be considered as a supercooled liquid
I love deadlines- I like the whooshing sound they make as they fly by.