I'm afraid I must apologize; I can't comment on these stories because I don't know what they are really about. Here are some things they might be about:
- playing scrabble with cavemen (like this?)
- using computer models to travel back in time and talk to William the Conqueror (instead of using, say, an Old French dictionary)
- how the oldest English words are two, three, five, who and I (but not four, bafflingly), which have remained unchanged in sound and meaning for tens of thousands of years, in fact these words would be understood in Ice Age Europe! But then why are the cognates of five (fünf, cinq, coig, pāñc, pjat', pénte1) so different from each other? Or the cognates of who (wer, qui, cia, kaun, kto, poi1)?
- how the words dirty, squeeze and guts will be the next words to disappear from English - so use them now while you can!
The Log tells us that the research behind this media frenzy is Frequency of word-use predicts rates of lexical evolution throughout Indo-European history, in Nature 2007. But as presented in the news, it has nothing to do with Proto-Indo-European, at least as I understand it. The time depths given - 15,000 years, 20,000 years, 40,000 years - are way older than the estimate for the breakup of Proto-Indo-European, which Fortson suggests is 2500 BC.
1. German, French, Scots Gaelic, Hindi, Russian, Ancient Greek