With some mounts the final slew misses by a small amount and repeated platesolves/syncs/slews do not improve the accuracy with more tries. Also, a slew is also a fairly extreme move to do when you are close.
Instead of slewing, use slow motion in ra/dec to do the final centering - perhaps when the distance is less than 10’ or something. There may be backlash in dec. that won’t make it predictable exactly - but it should converge to a given tolerance - unlike slews.
Frank
Alternatively - use slew but correct for the measured error when slewing to a given RA/Dec value. For some mounts this error will be repeatable and if you measure it, you can correct for it. You just set the target to be a bit off from the actual target.
If you did this, it would remove the need to sync at all - because this would correct both for the sky model of the mount - and for slew error.
And it should converge on the final destination - unlike repeated slews/syncs.
Frank