I'm not saying QM's mechanisms are the same boxing marbles. I'm just saying that separating entangle particles isn't just more interesting than boxing marbles, IE, all the modifications that are going interfere with each can be seen as already having happened - which is why QM doesn't allow you to transmit information faster than light.
Fair enough, if it's just an analogy I agree it gives the right intuition.
To be clear, what really happened under MWI is that the wavefunction "branched" into two around the time you created the two particles, and the two branches spearated from each other in configuration space just as the two marbles in your analogy separated in physical space. When you did the measurement, you merely realized which branch you are in.
Where the analogy fails is that the probabilities in QM are inconsistent with there actually having been some hidden choice of marbles made beforehand.
