Why PubMed is so important in the NIH mandate

Some us of know the following phrase by heart:

all investigators funded by the NIH submit or have submitted for them to the National Library of Medicine's PubMed Central an electronic version of their final, peer-reviewed manuscripts upon acceptance for publication to be made publicly available no later than 12 months after the official date of publication:

PMR: Why is the "PubMed" bit so important? Why wouldn't an institutional repository do? Or a departmenal web page? After all, Google will index these and find them, won't it?
I realise that the following is going to upset some people but I have to tell it as it is….
The reason that PubMed Central is so important to me is  thatit is not"just an archive", not "just an enforcement agency", not "just a repository". It is a living instrument of scientific research in a way that the others are not – and won't be. And it's evolving rapidly. It will, I predict, overtake commercial abstracting organisations (you know which ones) during the next decade. It will become the single most important scientific information resource in the chemical/biological arena. Together with arXiv and SCOAP in physics, it will be part of the twenty-first scientific information enlightment.
Why do I say this? Because for the last 30? years the NIH has had first calss teams working on the modern practices of informatics. That's shown by Pubchem, which is now seen by many (not yet all, but most of the enlightened ones) as the first place to look for chemical information on molecules. Why? First because it's Open, but more importantly because it has a modern approach to information. Pubchem will continue to overtake the conventional molecular databases. That's why there was such a fuss when Pubchem was launched.