The StarAdventure Pro is a camera tracker designed for DSLR cameras and lenses. They are great with medium length lenses (50-200) or something like a Redcat at 200mm. That mount is really not designed or suitable for a 6" newt with a 1200mm focal length. Too much weight and too much focal length.
Your photo of M33 is consistent with images produced in EAA with these mounts. The only way you are going to get your individual lights to have nice tight round stars with your mount is to decrease exposure time until they are round. This might mean you need to be shooting at less than one second. Then you are forced to crank the ASA up to the point that the noise will simply overwhelm any signal you may have. If you can't upgrade your mount at this time, I think you would be well served to pick up one of the cameras currently popular for EAA like the ASI224.
The periodic error and mechanicals of the mount mount make a huge difference in arc second pointing accuracy. A little too much backlash or a simply a gear that was machined badly and a hundred other mechanical details done poorly means it simply will not work for AP. This is really not surprising. A respectable mount for this size scope is easily three times the cost of that package. Your scope and mount were primarily designed around price and to use visually.
Something else that will affect how accurately your mount points is how well you do your star alignment. When I wanted my LX200 to point well, I used an eyepiece with a lighted reticule during that alignment. Even then I would sometimes have to repeat that alignment a few times. A tedious and annoying job at best, at least it was for me.
Then there are other issues like focus, collimation, and coma that all have to be solved. Perhaps you have most of these items addressed as you stated, but the proof is in the lights and I don't simply don't see it in yours. I am sorry if I am being overly blunt.
I am not sure you are aware of this, but there are differences in apparent motion of the stars including rotation based on where in the sky you are pointing. This can be used to increase exposure time without star trailing. There are threads you can find here on the topic.
If you are stacking the lights that I inspected, your integration will simply not have the fine detail you are asking for. When I am integrating my lights I toss out the worst 20-30% and I apply a weighed average to stack what is left. You need to toss every light with noticeable star trails. This means you need to inspect them at a pixel level scale or use software that can do that for you. Averaging the poor subs will only degrade your end result.
Edited by idclimber, 29 September 2020 - 01:57 PM.