I think anyone who finds themselves fully agreeing with this article (in particular the assertion that teachers "can not abolish human difference") owes it to themselves to read MindSet by Carol Dweck (or at least familiarize themselves with her research; I actually didn't like the book that much). She argues that in almost all cases, initial differences in intelligence among children can be virtually erased by fostering a "growth" rather than "fixed" mindset (definitions here: http://en.wikipedia.org/wiki/Mindset_(book) )
The basic idea is that while we do have some amount of inborn talent (intelligence say), this is merely a starting point and can (easily) be improved upon. The problem is, we implicitly teach kids the fixed mindset (that their level of talent/intelligence can't be changed) when they're young, and as a result they sort themselves neatly into tiers based on their initial level of intelligence, and then stay that way. In this manner teaching the fixed mindset seems to be a self-fulfilling prophecy of sorts - we tell kids they can't get any smarter than they are, and as a result they don't even try.
Thus the mindset theory explains the phenomenon of "dumb/smart kids growing up into dumb/smart adults" equally well as assuming that fixed talent actually exists. Moreover, research has shown that adopting the growth mindset causes increases in performance, productivity, grades, etc. - in short, it actually makes people smarter. This works whether you do it subconsciously (ie, priming experiment subjects with growth mindset ideas) or consciously (ie, actually telling people about the mindset concept and pushing them towards a growth mindset). There are a lot of really dramatic results given in the book that can't be explained at all assuming fixed levels of talent.
Any pointers to those bits of her research that you favour?
Post by fellow LW reader Razib Khan, who many here probably know from the gnxp site or perhaps from his debate with Eliezer.