Jeff P Jeff, I saw this on the FB forum the other day and played around with the S30 Pro I have during the daylight. I can think of 3 points at which backlash might be a problem. When the mount first slews to a new target, when the weight of the Seestar shifts from one side of the meridian to the other when imaging through the meridian, and, most problematically, after a dither.
In the first case ZWO must have some sort of backlash compensation which is pretty standard practice. But I am sure there are variations in the amount of play in the gears from one unit to another due to manufacturing tolerances. Perhaps this is behind what some people experience when the Seestar rejects a bunch of frames at the beginning before settling down or, even, requiring the stacking to be stopped and then re-started. A weight like you show should help as your experience shows.
The second time when backlash should be a problem is when the scope's weight shifts from one side of the meridian to the other. Playing around with the Seestar in daylight I could see this happen quite clearly. It occur slightly before or slightly after the mount crosses the meridian depending upon where the declination axis is pointed as this determines when the weight shifts from one side of the meridian to the other. Here there would be no backlash compensation if one is still imaging the same target. With the additional external weight I think one would still experience an issue as the mount crosses the point where the weight shifts from one side to the other. As it shifts, the motor gear has to catch up with the gear driving the mount so I would expect some initial frames still having a problem.
The most problematic place where backlash may be causing issues is after a dither. Again, ZWO should have some backlash compensation applied but also again, there will be variation from unit to unit and some may have more problems than others. This would be the worst case scenario since dithering occurs after every 5 frames. I have looked at some of my own data and it does appear to me that I see most dropped frames at or right after the dithering frame. I need to check more data and more carefully before I make this conclusion, but it does seem quite likely to me. But the counter argument is the fact that I, and many others can get very high capture rates, even approached 100% on some targets even though dithering is still taking place. I would have to reconcile that before being able to conclude that backlash after dither is a reason some are having more problems with dropped frames than others.