Have never heard any given number such as 1 million hits. Do you have access to a Renishaw MCG? If you do, you should have statistical data to show how your measurements have been running. If you don't have access to one simply write a looped program for your sphere and use min/max to verify your individual runs. If your TP200 is giving erroneous readings you will have a high range for a variation of over 4 or 5 microns. I ran a TP200 on a Bright Apex 707 and never varied more than 3 microns measuring at 25% measuring speed, volumetric accuracy should be below 10 microns, spec on the 707 are 13.8 microns.

Duane Wade