You are pretty much correct.

The causal reason that Bayes' theorem is in the article is because Luke wrote that part before I was involved. If I had written it from scratch, I probably wouldn't have included it.

But I think it's reasonable to include it because this is an article for laymen, and Bayes' theorem is a really important piece to have your knowledge of rationality built around. Furthermore, the discovery of Solomonoff induction was motivated by the search for objective priors, which is a Bayes' theorem thing.

But you're exactly correct that hypotheses either match or they don't. The updating of probabilities occurs when you renormalize after eliminating hypotheses. Also, I've thought about it for a while and concluded that no modification is needed to represent probabilistic hypotheses. I may write a piece later about how Bayes' theorem is consistent with Solomonoff induction.