as far as I know, yep fitting the "raw score" of a test to a bell curve is exactly how it's done. And often the score is sort of "localized", for example only other scores from the same country and done in the same year are compared.

(one related example is the flynn effect)

IQ is in reality a very rough metric, I think the only widely accepted practical use is to detect developmental or mental issues (often associated with an IQ below 70), and even then you need to consider that eg. someone who never received adequate education may score lower than what they "should"