Doing raw data visualization and being very careful about not making conclusive comments seems like a proper way to step in the subject without pretending to know more than anyone else. Also he didnt cherry pick what would have been the most convenient ones to make a point.
Compare that with typical media science reporting which is often much worse and misleading at best when not grossly wrong.
Quality work starts with data exploration and data mining. I see this article as an invitation to check the actual data thats available, if anything.