Ahh, so we are on the same page then! No problem.
I think so. From my perspective, the QC part was obvious since (I thought years ago, and still do) that is most likely the best (if not only
) way to control the other variables. That made it a 'no-brainer' as regards a test scenario that's as uniform as possible.
The rest is just getting the tester to 'follow the script', make the same movements etc while avoiding anything not in the test....and last but not least, being willing to both recognize when there's an obvious variance that would affect outcome, and 'do-over' if appropriate.
So far, this has never been done to my knowledge; certainly not been done without the issue being observed, and/or in a test that wasn't compromised.