As a committed perspectivist, I agree with its primary (and modest) conclusion that exploratory tests can be more compelling than confirmatory tests.
That is, there exist conditions under which we should take an exploratory outcome more seriously than a confirmatory outcome.