It’s actually that the quadrants are meaningless and are the result of bad charting - you can see the publication linked in the article. You get the same results from random data as DK got because they are charting a data relationship not results.