The Attraction of Complexity

I can't follow the article's conclusion at all.

The data is very clear that there is less change/expression on complex code. Aggregating it by function seems like a bad move - if those functions were broken down into pieces, you would end up with more simple code, that from the simple model from the article would require much more change than the complex function.

This analysis seems to be based on some notion of "Cognitive Complexity", but I can't find its definition in the article. Am I missing something?