That is not his premise, your summary appears a bit faulted; whereas 1. He says if there is an absolute reality then evolution could not select us to perceive reality as it is because 2. When running evolutionary game theory algorithms he proved that a species would not be able to compete with anything less than “selection for fitness reality” showing that if there was any underlying reality beyond fitness it simply would not be possible to conceptualize.
There is no philosophical assumption in that particular claim of his.
“Idealism” as a concept in western philosophy specifically philosophy of mind is likely not the best way to frame his viewpoint, if one wants to apply some analytical thinking. Hoffman has quite simply applied a scientific argument that appears falsifiable to what is known in the east as “non-duality”.
This (smaller) school of eastern philosophy has a surprisingly strong analytical tradition that has evolved well over two thousand years with an uninterrupted transmission of what are known as the “two truths” doctrine, the doctrine of absolute and relative truth.
Remarkably, eastern “non-dual” traditions (matter emerges from absolute and relative mind) would be monists, not dualists, which unites them with western physicalism, which is also monist, or “non-dual” as they would say in the east.
By relying on “idealism” to frame the view I worry creates a false impression of the underlying scientific argument as opposed to the philosophical.
The assumption that matter emerges from mind (1 substance) or the assumption that mind emerges from 1 substance (matter) are both “nondualism” yet both of them cannot possibly be true.
Physical nondualism, the eastern nondualists would say, is just a step in the right direction in the discovery of the two truths.