My take as someone who has not read the original study: I am cautious about this for several reasons.
-- A person's genes do not tell you what language they're speaking.
-- University alumni magazines love to talk up research from their institutions and make its results sound much more 'decisive' than they are.
-- That they get wrong simple things like when Europeans noticed that Sanskrit, Latin and Greek were related by well over a century does not inspire confidence. (But hey, scientists don't need to read early 17th-c sources in Latin, right:-)?)
That said, if further study of the Nature article corroborated these findings, I'd be more than delighted. I'm definitely going to take a look.
--Antonia