Right, theory. And the theory is not supported by actual data, just by math and their own estimations.
So again, I point out that what they have is a theory, and not proof. His article is unpublished and therefore not peer reviewed. It's educated guessing