I fully agree that with any equatorially mounted telescope, including the Seestars, it is absolutely not necessary to have a level mount to get accurate polar alignment. I also agree that it is helpful to level the mount as it takes out the cross-talk between the Az and Alt adjustment axes, but other than making the PA procedure faster, it doesn't make it any more accurate.
I also agree that there are other factors at play as to why some folks have better success with minimizing star trails and getting less dropped frames at longer, i.e. 30sec, sub-exposures. Maybe as some suggest, gear backlash comes into play more for some than others. But there are three other factors that I think are often overlooked which can explain one one person can get much better results than another person or why one person can get better results sometimes and not at other times. Both are due to apparent field rotation.
While we do our best to get an accurate Polar Alignment, we are never perfect which means that the apparent field rotation of the stars comes into play, more or less. Since field rotation goes as the cos(latitude) x cos (object azimuth) / cos( object altitude) folks at different latitudes with the same accuracy of PA and looking at the same object will see more or less field rotation. Unfortunately, when people post about their results, good or bad, they don't share their latitude information. I am at latitude 38 degrees so I will see 80% more field rotation than someone in Ottawa Canada at a latitude of 45 degrees. Also, if I image objects while they are at azimuth angles closer to 90 degrees than 0 or 180 degrees I will experience less apparent field rotation. Similarly with object altitude. So this makes it difficult to compare notes from one person to another unless they share more specifics than just whether or not they level their tripod, add weights, etc.
Finally, I suspect that inaccuracies in the internal compass and accelerometer come into play and vary from one person to another. The way the EQ mode works, after the plate solution is complete we get feedback on our alt and az angle offsets. I assume, and ZWO tech support and chime in here as needed, that this is coming only from the plate solution and not the internal sensors. However, as we adjust the az and alt to minimize the deviation, the internal sensors are now giving us feedback to let us know if we are getting closer to 0 offset or farther away. Any inaccuracies in these sensors, whether because of poor calibration, fundamental imitations, or variance from one Seestar to another, can also play a significant role in how accurately one can PA the mount. This is where ZWO tech support should chime in and give more details on both the accuracy of these internal sensors so we don't go chasing our tails elsewhere.
