Seems like you guys know what your doing!!!
I've made elipticals in our shop, encluding edging at 45 deg. and,
1)Multiple parts, or lots of spacer material is important. The more symetrical the layout ( in terms of distribution of surface area), the easier to achieve 'test'. On things that really need to be flat (1/20th wave), the real parts should be in the middle so to speak, as this avoids roll off problems, and a single part in the center is not the best situation either.
2)Consistant material type is very important. Harder or softer materials polish at different rates, making it impossible to control test.
3)Parts should be blocked on very flat tooling (usually pyrex) using low stress blocking pitch, and sloooow cooled to prevent stress from building up in the material which will cause springing when deblocked.
4)Before final blocking, both surfaces should be 'flashed polished', as this removes alot of the stress still present even after fine annealing.
5)If springing still occurs, compensate with the on block test, or there are even more exotic blocking procedures....
I'd better stop here
Bill Marriott Forest Knolls, Ca. btk@ix.netcom.com