The title of the article is misleading. One may think that R is a tool is worse than other alternatives. It's more like "current methods (or simple methods) are not enough". R is as good for data science as any other tool (and arguably is even better thanks to its extensive library of packages).