I also wonder how different this approval is in terms of meaningful scrutiny from a normal approval.
FDA approval is for the specific test, not for the machine itself (as those did not need to change).
I was searching for analogous situation, where these machines were able to accurately detect other infections and it seems that they were able to perform at 100% for flu-A at 100% accuracy. [2]
My impression sofar, is that FDA was definitely asked to work faster and more efficient, but it does not seem like that they had 'cut corners' so to speak, and approved something out of the 'blue'.
[1]https://www.lucianne.com/2020/03/21/rapid_coronavirus_test_a...
[2]https://www.labmedica.com/microbiology/articles/294780980/po...
So this doen't seem to be some speculative thing. What's different from their other tests is the RNA target. And that's public knowledge. So it's arguably just procedural.