It seems like there isn't a day that goes by without another 5-axis problem on the VF2!
My boss worked on a set of parts that needed smooth blending on a cylindrical shape and had some slight offset issues.
It seems that the centerline of the a-axis isn't quite aligned with the midpoint of the b-axis table and may have also caused some of the previous alignment issues. We were able to find out if it was off by cutting a flat on the block with the a-axis at 90deg, then rotating the b-axis 180deg and cutting the same flat on the other side, measuring it. It seems to be a whole 8 thou off (4 thou per side), which is quite large and I'm shocked that the issue hasn't really come up too much before.
Thankfully to find the true center was as simple as getting the height offset from the b-table, which is done by edge finding the top of the b-table when the a-axis is at 90deg, then flipping the a-axis 180deg and finding the opposite edge, then subtracting the difference and dividing it in half. It seems that this solved the issue though I didn't have enough time to do more extensive testing to ensure it all worked properly.
My boss thinks these 5-axis issues were coming from when it got re-sealed, and some of the calibration systems threw off previous numbers, so we may need to do a more thorough look-through of all the offsets.