sshyang - I think you missunderstand me. I have extremely good reliability with SGP (V2.5.1.17), so much so that I have too much data to process! I have no issues with meridian flip and when I'm using slew/center with the PMX it works perfectly. In my other post, I was prematurely trying out the new beta program with its sync offset feature.
At the moment I have a TPoint model - so I do not use 'center' Currently the recovery process aborts at the meridian. It is clearly documented that it does. Obviously, if the system is in recovery mode, it is normally the case that there is cloud cover and so any centering process would not work, My suggestion was that, in the case of recovery mode, if it gets to the meridian, rather than aborting the sequence, it executes the meridian flip but does not do any centering. (In fact if centering had been enabled, it could be executed just before the sequence resumed.)
The existing features are stable. There are a few minor concerns, but what software is perfect? The software is, however, part of a complex system, which includes the user, which is not under their control !