The trouble is that the same empirical data is also consistent with the theory that when efficiency improves it attracts market interference to capture the surplus.

And then it could be either one, or some combination of both, but the people doing the capturing prefer to direct attention to an alternative explanation so they can continue their extraction.