Earlier quoted context omitted.
That is exactly what he argues. But I think it's a tough position, given that relational quantum mechanics also works "out of the box" and doesn't require ontological realism and a deeply speculative metaphysics. Sean Carroll discusses this with Carlo Rovelli in two separate Mindscape episodes.
I don't grasp essential the difference between everettian and RQM. I'm both cases different observers observe different results. IIUC the "worlds" on the many worlds interpretation is just a convenient way to talk about the relationship between an observer and the possible states it can observe after the system and the observer have been entangled. The emergence of a world branch is not something that happens at some…
I am sympathetic to the idea that no one interpretation is 100% correct and that we shouldn't be deeply attached to any of them. Each may have its usefulness in different contexts. Independent of that, some interpretations more accurately describe the quantum world than others, surely.