My guesses/observations would include: Twitter r uk email address list eally is cannibalizing the web's link graph, or at least, the blogosphere's and Google seems to be using Tweet counts in some way (though possibly only in the QDF algo). The acceleration rate of link acquisition and the freshness of new links is having a more dramatic impact than before, and the "old crusty links" paradigm may be fading a bit. Brand mentions and keyword associations with brand names are influencing the rankings more and more.
Un-trustworhty link patterns are conferring more filters and penalties than ever before. QDD is as strong as ever, and vertical results are more prominent than at any time in the engines' histories. Google and Microsoft both know more about traffic and surfing habits than ever before, and this data is likely being used to, at the least, quality control for potential algorithmic misses. Ad blindness is worse than ever (16% of Internet users are responsible for 85% of all ad clicks on the web), forcing the engines to make ads more relevant and more obvious to continue earning revenue.