One of the issues I have with the MWI multiverse theory is infinite state proliferation. Let's take some starting assumptions - reality begins with a single state, it progresses in a quantized fashion (it "ticks"), and the state change duration (the tick length) is equivalent to Planck time.
After a single tick, assuming every possible variant is explored (every "choice" that can happen does happen), you have three options, which really resolve to two possible subsequent potential states. Either the exploration of states collapses back to a single state, or every single possible state continues as a new existent reality (effectively an infinite number of new states), or a subset of that infinite set continues. Without any way to quantify the effective subset to some finite subset, that really leaves two options - one new reality, or an infinite number of new realities.
Bear in mind, that's after one tick - and you get roughly 10 ^ 44 of those every second. So the complexity of the possible choice sets (if the multiverse does not collapse to a single state after each tick) is a proliferation of an infinite number of choices for each starting reality, every tick, again and again.
If an infinite number of realities boggles the mind, how much harder is it to comprehend the vastness of allowing all potential realities to propagate, as would be required if the set of possible states did not collapse each time the state progressed?
And if the set does collapse back to one reality, what dictates the choice, and surely that precludes continued exploration of each of those infinite choices that have just collapsed, ruling out the opportunity of existing or exploring the multiverse (hopping realities, that sort of thing).
If one assumes that the complexity of the reality state is not actually infinite, but is in fact bounded by some finite but hideously complex state (as an arbitrary example, "counting up" all the available particles and extrapolating a finite state complexity from that) then that would suggest a comprehensible non-infinite set of reality choices, even though the complexity does still scale at a nearly mind-boggling rate.
However, if that's the case, then even if the multiverse is comprised of every possible alternative state that could have ever existed - it's still not infinite. That means that it's possible to rule out some hypothetical states from the actual set of available multiverse states, assuming that one states with an infinite number of possible hypothetical states.
If one takes current computational parallelism as a paradigm, it could be hypothesised that "running" the multiverse equates to execution of all available reality states each state change transition, followed by selection of the valid state. Parallels can be drawn with our current understanding of quantum computing although I wonder if the execution is speculative (leading to a collapse of the reality set) or continued. If the Copenhagen interpretation is to be believed, the observer is the fundamental cause of the collapse, but not given credit for the result of the collapse. However, given that the observer cannot be treated classically but any observation should be regarded in itself as a quantum phenomena, what does that mean in causal terms for the initial state split or propagation?
Just because I can't wrap my head around the vastness of it all doesn't mean it's true. Still, it makes my brain hurt.