By calling it a universal word, I guess I thought they were referring to the concept of linguistic universals , which are interesting since they might suggest something about the deeper way in which our brains work. To make a claim about linguistic universals like that I really think they should study more than 10 (31) languages.
According to this page , 44% of people speak an Indo-European language, and 96% speak a member of one of the top 10 language families. The remaining 4% includes 84 language families. I imagine there's a ton of variety there, and for understanding human language and cognition they all probably have just as much value as the larger language families.