I don't like Ed, but it seems that Luu spends far more attention documenting failed predictions than checking other predictions that might have been correct.