Mirror "B" tested as 1/4 wavelength.
Using Dave Rowe's
FringeXP
program and this interferogram here is the Test Report that resulted:
The RMS wavefront error and Strehl ratio are again much better than I predicted.