I have one thing I think you need to try and that is to go back to v1.03 of the HFG board firmware. That's the only way I could stop the jitters and fix the yaw coupling issue. I've tried every version since v1.03, and they all have this same set of issues. They broke something, starting with v1.5, and can't seem to find out what. Until they do, I would definitely use v1.03.
The first video below is with v1.7r2. You can clearly see there's a horizon drift problem, and a pretty bad jitter/shake problem. It rolls left, slowly and after landing you can see it level back out. The jitter shows up in the roll axis the worst.
In this second flight, the only change was to go back to v1.03. No horizon problem and no jitters.
George here will tell you they can't duplicate this problem, but personally, I just don't think they are trying hard enough. For example, in another thread, he said that for all their test setups, the board is mounted vertically, and maybe the problem is related to the fact that for our setups, we mount it horizontally. He implied this is a design problem on our part. That is a total crock. Why do we have to make design changes to test their
issues? Why can't they change their test setup, and just see if that is in fact a problem. This is just pure arrogance, in my opinion, and horrible customer support, at the very least.