Possibly you might have both a small crankshaft flange runnout, and a
prop out-of-balance situation.

Your measurements are probably in inches per second (ips). If so.
reversing the prop should not give that much change in the optimum
vibration compensation vector. That makes the prop balance as rebuilt
suspicious.

Nevertheless, there is probably an inherent engine imbalance too, which
could come from a number of things besides just flange runnout.

You mentioned both a 750 RPM and a 1500 RPM rough spot. Is the
airframe responding at 12.5 Hz and 25 Hz - or is the response always at
the same frequency? I realize it may be hard to tell the difference.

It isn't the governor, and if the vibration came on suddenly on prop
rebuild - not related to the crankshaft recall.