A hypothesis must grow from observation. By this definition, any testable hypothesis comes with evidence. (This is why a hypothesys can "transform" into a theory, it's a first attempt to explain facts and observations.)
Applying that to this situation, it appears that the hypothesis is indeed based on observations. As it is also testable, it makes for a scientific hypothesis.
That said, the criticisms leveled against it are not invalidated but this fact. At the moment, the observations are very loosey-goosey and the conclusions that lead to the hypothesis may have some bias, and an important question - correlation or causation - isn't addressed.
As I see it, anyhoo.