Few comments about Y dna and language correlation. We know that ancient America was populated by a very limited number of Y dna more than 15.000 years ago. Mostly few branches of Q1b. Yet the linguistic diversity of pre Columbian America is huge. Here an excerpt from Wikipedia.
Over a thousand indigenous languages are spoken by the Indigenous peoples of the Americas. These languages cannot all be demonstrated to be related to each other and are classified into a hundred or so language families (including a large number of language isolates), as well as a number of extinct languages that are unclassified due to a lack of data.
This diversity is impossible to explain by new migration. Only two new migrations are known to have occured later. One of which introduced Na-Dene languages ( famous Apaches ) with the Y dna C2. While the other introduced Eskimo- Aleut languages with Q1a? branch. Those two migrations occured less than 6000 years ago and some believe that Na-Dene still preserved a connection with a Siberian isolate known as Kets.
What this mean for linguistics and genetics? This mean that linguistic has its inherent limitations. After 10.000 year of separation two related languages accumulate so many random phonetic shifts that it is practically impossible to reconstruct their common vocabulary.
This is especially true if languages are illiterate and do not have developed religions and sacred languages. Are mobile hunters. A similar situation can be expected in Paleolithic West Eurasia. Though higher population density in Near East probably made languages more stable. But in any case even if two languages descend from G2 rich pop then some 15.000-20.000 years later it would be impossible to prove their genetic link.