That's fatuous. The ubiquity of polysemy across languages is attributable to the stability of the organization of cerebral language centers across individuals. It's not necessary to posit an arbitrary parameter and then to create "data" purporting to demonstrate the applicability of that parameter in a derived statistical model. Everybody knows that!

Aloha,
Mark