I agree with you point theoretically, and I too struggle with the question of who or what maintains the simulation. I don't, however, view this as a reasonable criticism of the theory in itself. Just because we cannot fathom the origins of the simulated reality doesn't change the statistical argument that this is, in fact, a simulation.
On the other hand, it's so far removed that our notion of computing power might not even apply to the real base reality. Our notion of physics might not even apply.