Joking aside, the authors of these studies have highlighted the low statistical power of research in a number of fields (see also http://www.nature.com/nrn/journal/v14/n5/abs/nrn3475.html), which is a serious issue that often leads to over-interpretation. It's as if researches willing to publish more are rushing hasty studies out the door.
It's a good thing that people can measure these things and ring the alarm
Depends. By and large, Ioannidis's paper is pointing out the basic consequences of the alpha and beta parameters in Neyman-Pearson hypothesis testing, so there's a very good argument that this paper is merely statistics, and statistics is generally considered to be mathematics, and mathematics not science but its own thing.