I read this thread and the cloudy night forum with great interest. It's quite a revolutionary approach to guiding. I have to join the skeptics though
We look at the sky and we guide accordingly in reference to what we're imaging right? Putting aside what Peter already pointed out that we have orthogonality problems, wind, flop, flexure, etc... you name it: still if we had resolved all of the above it would still be like being bolted solid to a boat deck and taking pictures of the sky right? I mean, you can't compensate only one RA rate when there are so many other variables at work?
Sounds just like common sense to me to use the stars as a reference?