Interesting. But there's one thing that makes me very uneasy about these kinds of studies. To make performance comparable they have to radically reduce the complexity of the tasks on which they measure. It may well be that variance is much greater between individuals for average real world tasks that allow for reflection, communication, etc.