Musk literally said "we're going to dumb things" and keep the things that aren't dumb.
This "going to do dumb things" is completely on brand for him with his 5 step manufacturing improvement process, step 1 of which is "make your requirements less dumb".
This is actually how Nike works, having worked there. They try all kinds of weird things aren't aren't afraid to, and then drop the ones that don't work with no regrets, keeping what works. Try new things, fail fast isn't a bad strategy for innovation.
Whether or not it'll work out for twitter remains to be seen. Especially with the rest of biased tech still upset that they lost their monopoly on the narrative arrayed against him.