Had 4 great nights in January and 3 of them had FWHM values around 2-4 which I think is probably normal UK maritime skies.
However a fourth night which looked visually good gave values around 6-8. Aperture was adjusted accordingly.
The full moon gradually rose, but I think this would only affect sky (and star) readings. The moon of course helped highlight any hazy thin cloud.
The surprising thing is the errors were only slightly worse.
I wondered if there is a limit at which you quit observing. I'm guessing no simple answer to this simple question.
Maybe others have a working rule of thumb to guide at the telescope. Unfortunately my imager does not offer FWHM at the scope.