Hi Steve,
I've found this topic a few minutes ago.
As Andres wrote, the astrometry was "re-invented" with the latest updates of PixInsight. The new methods exposed with ImageWindow object are today, what I was looking for since long. Thanks to Andres and the team.
In some cases, distortion can be a problem. When you see star centers not matching the mask, tweek your ImageSolver with higher magnitudes (i.e 14..17). This takes more time but should give you better results.
If you have doubts on non-masked stars, please verify the positions with Simbad and send me the coordinates.
Regards
Hartmut