I could not really believe the 521 year thing, so I had to look it up. Its 521 years for a 242 bp fragment (even nature news reports it like you though) which e

I could not really believe the 521 year thing, so I had to look it up. Its 521 years for a 242 bp fragment (even nature news reports it like you though) which ends up at a degradation rate of 5.50 × 10–6 per year per base pair (and they estimate double that for nuclear DNA versus the mitochondrial DNA in the study). Still much too short for anything in the million year range though.