I have mentioned this request but I don’t think I entered it as a feature request.
My cge-pro does not land exactly where it is told to - but it does reliably land a bit off from the target. This means that repeated solves/syncs/slews will not converget to the target and I can’t center reliably. I need to do it manually - particularly to place a guidestar on oag.
Other mounts are not designed to sync all over the place - so for them a sync should be avoided.
And other mounts seem to have similar unknown problems that cause them not to converge to the target after multiple iterations.
I think many of these problems would go away by not doing a sync at all - but instead have SGP keep track of the error when converging on a target - and keep adjusting that error on repeated solves/slews. Essentially sgp would calculate its own sync offset when centering a target. There would be no need to keep track of this offset because whenever you start the centering process you set it to zero - and it only exists as you converge on the target during the centering process iterations.
This would mean the initial slew to a target might be a bit off since it doesn’t benefit from a sync - but at the same time if you sync in one part of the sky and then go to another - the sync may be detrimental to accuracy and you would want to unsync before you did the long slew. Additionally - if someone does want to sync - they can always do it manually near the target.
The basic idea is to have an error vector E, in ra/dec, that is initially (0,0). You slew to the target, T, plus the error, E: SlewTo(T+E). Then you solve and find the new error of the target from the new position, P: dE=T-P. Then you adjust E with E = E + dE. Then you repeat and slew to T+E.
As long as the slews are repeatable this should work. And it avoids sync, and should handle many unknown errors that currently prevent mounts from centering accurately.