I also have a similar problem with meridian flips as Frank. I didn't use it during spring/summer, but since late summer I've had a heavier Newton telescope for which I wanted to use the original mount tripod. The meridian flip failed twice in September. I noticed that after the failed flip, the telescope did not return to the park position, but landed significantly away from the meridian.

I have since decided my imaging sessions, one before and a second after the meridian transit.
Since this procedure is time-consuming, I made several attempts last night to understand my problem:

1. PAA was successfully completed, then the telescope returned to the park position (time 19:33:28), RA: "02h 07m 54s" DE: "90° 00' 00"
2. I then approached several stars and then returned to the park position because the solver partially failed (time 19:49:25 (solver OK), 19:54:03 (solver failed), 19:58: 37 (solver OK), 20:29:01 (solver failed), 20:31:34 (solver failed)). The parking position was different in each case, but DEC was DE: "78° 26' 40". Shouldn't the telescope have returned to DEC 90° 00' 00''?
3. I did another PAA after that to correct the mount model. The overnight session with the scheduler after the transit was then successful. The session was canceled at 05:41:27 due to dusk.
4. This morning I manually put the telescope in the park position at 07:44:09, DE: " 90° 00' 00 (!!) , because after the scheduler was aborted, the park position was not approached and the focuser was not in the 0 position (can one do this with a script?).

I loaded the “failed-align-frames” and the log to www.dropbox.com/scl/fi/61fbrglby93reqgv1...fd2cm2viyn531kx&dl=0 . I never used dropbox so far, so if there is any issue with the link let me know.
Any information about my issue is welcome.