I'm just now reading this post and am mystified by the quote you chose to pull, Bostonian. What this paper actually shows is good results for something I've long thought should be done: universal screening for giftedness. When the district began screening universally, they found way more traditionally underidientified students, whose performance in school improved. The only downside was cost--but consider the cost of NOT identifying these talented students.
I live in a district with a similar method of ID to the "old" method described here and as time has gone on, I have come to learn that MANY of the children in my kids' program (which has a hard score cutoff) were retested privately after not quite making the needed score when tested by the district. This comes at quite a cost. Guess who doesn't bother with that? It's a bad system, especially because initial ID can be done by parents...
