We don't know: * how many new short positions were opened * how many old ones were closed * what happened next to the shares involved in those transactions where retail investors moved 75M shares * where the shares came from in the trades where retail investors bought 74M shares
That's why Levine discusses many possible stories behind those numbers. The only real thing that many retail sales really does is contradict the "we're all HOLDING!!" story.
It could be that very little of that was closing the original Melvin positions... but we don't have evidence, do we? How can you prove that retail didn't sell 75M to people closing short positions, and then retail bought 74M back from the brokers those shares were returned to, while meanwhile, a bunch of new investors went short because 300 bucks is a bonkers price and they're betting they can wait it out a few months? Or anywhere in between?