Fascinating to hear the proposed engineering solutions. I've found the reality with my chinese mounts (particularly with the Newt on) is that balance after a flip and tracking quality also depends on the DEC you're point at. If I'm going to leave the rig running automated with a few flips possible I'll balance it neutral and accept the possibility of some tracking drift. Once you start doing automated runs over a few nights an unexpected benefit is you get so much data you can afford to throw out more problem frames and the mathematics of integration means you gradually throw out pixels due to poor tracking. Sort of a "brute force" mathematical approach to increase the probability of tightly imaging a given piece of object detail.
I've been mulling over these things a bit since Ivo gave his great talk at the AIC last year. For those of us unable to get into quality mounts at the moment, you can still do a lot to lean things in your favour if you understand the limitations of your rig and how to work with it. Automated runs for much more data definitely one way of doing it. Heck, I treat any data collected after I'm in bed as a bonus anyway
