Brilliant - this is perhaps one of the clearest explanations of the disagreement between each of these interpretations I've seen, and it makes crystal clear why we can't hope to observe any differences: the questions are all fundamentally "why" questions, which we can almost never make direct observations to resolve.

At the root, we're always asking why some random variable ended up in a certain state. Our "interpretation" just changes where we ram that annoying random variable into our theory, which largely depends on where we're most philosophically comfortable with uncertainty showing up.