Hi Matt,
I tend to agree with you that math is efficient because we have developed it for areas in which it is efficient, so it's kind of a pointless question if you interpret it literally. It's like asking why email is efficient at delivering messages.
However, I think you're missing the more relevant question, that is, are there aspects for which math cannot be used, or it is "efficient" for everything? I see you're not very convinced by the mathematical universe hypothesis. Arguably you are right that saying that there are more mathematical structures to be discovered is hardly a useful prediction, but the question is there a mathematical structure for ANY observation that we can make (natural science or any other), or not? Maybe you have a chance to look at my essay, I think you might find we have some things in common.
-- Sophia