That's too bad, then. A good paper should not be retracted simply because of so many misinterpreting what it's talking about and creating some pointless controversy about it. The underlying "phenomenon and prediction" (namely, the bimodal distribution of proficiency in basic CS) stands up to scrutiny.
