This seems to be somewhat biased in favor of vocabulary. It would be interesting to see if there are similar methods for estimating breadth of vocabulary from 1

This seems to be somewhat biased in favor of vocabulary. It would be interesting to see if there are similar methods for estimating breadth of vocabulary from 10 words and checking for correlation. You could perhaps use this to scale the score. The cynic in me feels like this would just result in headlines like "Knowing more words makes you creative"