The study is incredibly misleading. It clearly argues that mid-priced gas cars are cheaper to operate than mid-priced EVs. Not in extreme cases, or some cases - there are no qualifiers. This is an example of how statistics can be manipulated to say pretty much anything you want.