so wait, CGscribe, do you agree that chart is prolly what they use on the real USMLE?
I really think %correct alone is not perhaps as important as is %correct relative to other testers. My guess is that each question in the bank has a %correct and at the end of the block it calculates an average. It then compares your score to the avg and repeats the process for all the blocks. Through a lot of statistical play, I bet they find the mean/avg for your test= national 3-digit score (~220). And then they take ur average and see where you would fall on the bell curve.
This just a thought, but i don't think it's a set in stone that X%correct=xyz since no one prolly gets the same test. Not to mention they may even factor difficulty of experimental questions since students may potentially waste more time on them. I think it fundamentally beyond our understanding really, haha.