I have been in this aDNA trade for some decades now, and I have seen a number of miraculous techniques and substances being introduced, advocated for, awed, till the enthusiasm fades away. Some of them proved to be useful enough to get adopted into the general aDNA working-flow. Such was the case for sodium hypochlorite, that was advocated as the solution to all contamination problems at the 2nd aDNA meeting in the early 90ies in Washington. The silica binding-technique, that was the hottest thing in aDNA in the mid-90ies, also proved good enough to get included into most extraction-methods. Others, like PTB and amino acid racemization, are since long forgotten. But one thing they all hade in common was that none of them managed to live up to the expectations created by the super-hype when they were announced.
Having seen all them different super-ingredients fail, or at least not preform the kind of miracles they were supposed to but only being useful, I have grown sceptic to anything that is said to have the potential to revolutionize aDNA. That is the main reason for me having been hesitant to the superiority of using petrous bones for sampling. But statistics is something I do trust, and Henrik Hansen et al. seems to have done the kind of study that is needed to convince at least me. The petrous bone does indeed seem to be the most suitable skeletal part for ancient DNA sampling.