I've heard he selftypes ENTp.
You've heard wrongly.

Where he explores the 'system' more, creating what strikes me as a Ti 'system' for Reinin dichotomies, the theory, building blocks of it etc... eventually producing 667 dichotomies.
His argument is that something doesn't magically become a significant distinction just because you can give it a name. However, most of the new distinctions he suggests are much more indirect derivatives than the Reinin dichotomies.