How does a h1 link make sense? Shouldn't the h1 be the header of that particular page? IMO it cannot be logically the most important information on a page and at the same time point somewhere else? Or may I misunderstand your post?
To continue my post about what I call "weighted local PR": A lot of people refer only to the search term as locally relevant, but IMO the linkgraph of a website is much more important. Maybe the local weight is given to the entire website or just directories.
ATM it appears to me a local newspaper link is much more relevant than a dozen syndicated links from press releases from all over the world.
I also have made the observation that clicks may pay a significant role in rankings: I am watching search terms in google WT and see them rise in the serps to a certain point when they fall back and may rise a little bit again while other fall even deeper. The best search terms have also the lowest bounce rate in google analytics.
Matt Cutts was talking a lot of the new google infrastructure; well, to be able to compute the local relevance of search terms, linkgraphs, click & bounce there is certainly much more computer power needed then before.
It may also be possible that the SERPs are somehow shuffeled and polluted to test the algo improvements, to see if the algo is able to learn and to correct things based on (local) user behaviour.