I am still having the same height issues on the Haas minimill 4th axis.
My boss suggested I place the 4" gauge block up against the fixture plate and see what the height differences are. Hypothetically if all was correct, it should be within a few tenths (0.0001"). However, when I took a look at it, there was a noticeable height difference. This was good in the sense that it was a fixable issue, and I moved the design down in Fusion360 about 15 thou (0.015") and recut the top plane of the fixture plate. I re-ran the same program to bore the holes out and got the setup piece precisely the right height; I was pleased with this and thought my problems were over.
However, when I ran the next part, it came out ten thou higher than it should (ex. part should be 1.00" deep, came out 0.99"), which was almost identical to what I had before recutting the surface of the fixture plate. I was baffled as to why it's only ten thou off, and the previous wasn't twenty off. I re-ran the same program several times to see if the strain on the 4th axis was causing it to shift slightly, but to no avail, each pass was identical. I decided to just manually offset it by the amount it was off and get them finished up; this was no big deal and completed all the parts without any issue. I mainly wanted to figure out the cause of the problem more than the solution, which was easy.
While I was cleaning and deburring the parts, I realized that the bottom of the hole had a slight upward bevel to it. I asked my boss about it and took one look at the tool I was using and mentioned that the square endmills have an inward angle (which I was aware of on some level but didn't think about it in this context) them to help with chip evacuation. The endmill I was using was a 1/2," which I was using to cut a 3/4" hole (not actual diameter), which meant that the surface of the hole wouldn't be perfectly flat. The solution, of course, was to use a smaller tool, which I did. The strange thing was that I got no difference in the height; it was still ten thou off. I ran out of time by this point, so I had to finish up the parts and get them out of the way for the next set in the assembly. I hypothesize that the tool height differences are slightly off between the 1/2" and 1/4" being higher.