Much better article and discussion in the link that sctb posted. This article has a number of issues... it's conflating several different problems with fMRI without really explaining them, then arguing that it's worthless -- as opposed to just needing more groups to follow best practices.