To me as well, that was the most interesting part of the article. Almost sounds like communism :) but really it's more like the advice to not judge a book by it's cover. Giving everybody a chance is much more effective than funneling funds into a small "elite". As far as individuals and people are concerned at least.
Having said that, the model used in the research is way too simplified to draw conclusions from. "Talent" in it is just a change in the likelihood of a randomly good event or bad event. There is no activity, no searching for opportunities and no competition. They could substitute "talent" for electrical conductivity and events for being struck by lightning and the results would be exactly the same.