Here’s some discussion back in the day, partially related at least:
The real question is, what’s the impact on the end result? As mentioned, splitting the difference between focus steps give you a max potential distance of .0139” away from the actual point of focus. Using other internet info, a beam is considered in-focus with up to a 40% divergence in beam width. And the setup for the GF should have an in-focus (by the above standard) distance/depth of about 0.1”. So I’m not sure what the deviation would be for .0139”. I guess one could extrapolate it out and get a ballpark. Or make some small risers and test material at a few different heights around the focus steps.
And then you also have crumbtray variability which could be a couple to a few hundredths. 
And in the end, we aren’t even really measuring spot size, but kerf (a correlation between the two exists, so they aren’t independent, but they also aren’t fully dependent).
Either way, just thinking aloud more for the sake of thinking aloud than anything else.