What exactly do you mean by "power" here, if not the ability to predict real-world phenomena? In absence of it, what exactly would make it anything more than an exotic artform?
Like, consider: parabolas were pretty fully described by the ancient Greeks, purely as a symbolic abstraction. It was only 1500+ years later that anyone realized that they could also predict the motion of cannonballs and planets. But that discovery was completely orthogonal to the math - e.g. symbolic statements about parabolas didn't get any truer just because they now also described real-world phenomena. (And likewise when we later discovered that planetary motion isn't quite parabolic after all, that didn't affect our understanding of parabolas either.)
That's all I was saying here - that the "esoteric artform" part of math where one abstractly examines symbols is the essence of the thing, and the "predict real-world phenomena" aspect is a side effect that sometimes happens and sometimes doesn't.
You may argue that that, in itself, is powerful, in which case fair enough. But that "power" would be comparable to that of poetry or painting, which, in my opinion, does a disservice to the true power that mathematics holds. Mathematics is much more powerful than poetry and painting, because poetry never helped us build nuclear reactors.