Generally a good idea.
The other thing that can cause HFD to deteriorate (and is probably at least as common) is deteriorating seeing and the often attendant poor tracking plus “seeing caused atmospheric de-focus”. That is something that can’t be fixed short of laser star AO so I wonder if there is some way to eliminate seeing/tracking from the picture. In other words, how does one separate crappy focus on one’s system from crappy skies?
Maybe if the HFD has gone up, look at the RMS tracking error and if that has also gone up, then do not add another focus but if RMS tracking has not gone up (but HFD has) then do the focus. Since guiding is on a centroid, I suspect tracking would not be affected much by focus until it got way off but that should be looked at.
One could also compare the new (triggered) focus HFD to the previous HFD (that triggered the refocus) and if it did not improve, one could assume that it was not focus that was the issue and proceed accordingly.
OK, that does add a lot of complexity but I figured that the skies parameter should at least be pointed out.