Romer's examples aren't quite the same as what this page talks about. He describes papers where the maths is valid - the conclusions follow from the premises - but it's not tightly tied to the real world via words.