Like he says in the article, you can't define success so you can't really answer this question.
A more interesting test would be comparing randomly sampled students from standard schools against students from alternative schools of thought(Home schooled, private schooled, new age methologies etc.)
You would then have to find a way of generating tests or activities that creates meaningful comparisons of their abilities/attributes. The problem in the end, maybe that boiling everything down to a few attributes is rewarding people for gaming the system, versus changing the system.
A more interesting test would be comparing randomly sampled students from standard schools against students from alternative schools of thought(Home schooled, private schooled, new age methologies etc.)
You would then have to find a way of generating tests or activities that creates meaningful comparisons of their abilities/attributes. The problem in the end, maybe that boiling everything down to a few attributes is rewarding people for gaming the system, versus changing the system.