[tsimionescu]

On the contrary, it is still an open problem to rigorously derive the Born rule probabilities in MWI just from this perspective. Either way, the response will be equivalent to the Born rule: instead of positing that a measurement'outcome is proportional to the amplitude of the state, we posit that the number of branches in which an outcome happens is proportional to the amplitude of that outcome. Not sure why these are fundamentally different.

Also, the MWI idea of branching is no more satisfying or intuitive than the wave function collapse, which at least doesn't require an infinity of universes out of which some are much more probable than others.

Note also that there is only 1 of you in MWI, you just exist with different amplitudes in different states, but when interacting with another object, you become entangled with a single outcome and thus can no longer perceive the other states that other versions of you perceive. This is important, as otherwise physical quantities would not be properly conserved.

[GoblinSlayer]

The Born rule is statistical distribution, and statistics is certainly computable in MWI. It's a matter of converting the wave function into distribution basis, where it will have the Born rule distribution with amplitude close to 1, which means that observations have the Born rule distribution in most cases.