The only stats in the opposite perspective are "30%-50% of 90th percentile students regress" which seems like regression to the mean, and smaller gains for minorities and low income students at high achievement levels. That's not a super compelling counterpoint.