Since we both have the same issue seems likely it isn't operator error. I submitted a bug report on GitHub.

github.com/indilib/indi/issues/1932



If you can add any clarification or additional log file to help identify the issue it might speed solving the problem.

Thanks for doing this!

Jasem has implemented changes in the driver based on information received from Rainbow Astro about how the ASCOM driver establishes Pier Side. It is untested - I will test this weekend when I get home but in the meantime it is there to pull from GitHub and test for anyone.

Great news!
I am on stable update channel... unfortunately, won't be able to test until it's released.

The driver update now in GitHub correctly reports pier side and is able to reuse guide calibration over a meridian flip.

I installed INDI 2.0.4 and tested RST-135 with simulators.
Yes, it does report the pier side correctly and the calibration can be reused.
However, after the flip the guiding runs away really badly no matter if Reverse DEC option is checked or not.
After re-calibration the guiding is fine.
Still, something is not right.
I'll test it under the real stars in a day or two.

I was able to guide after meridian flip using PhD2 - I ran the tool to check meridian flip and it told me to check the reverse DEC box and with that it was OK - at least for one night with a single meridian flip so not extensive test.

Just tested with a real setup under the real sky and it worked!
Guiding could be reused and was fine after the flip.
Strangely, with the Reverse DEC option disabled.
I am using the internal guider not PHD2.
Hooray!

Thanks Jasem and Team!