Neil beat me to posting on this, but it actually turned out more interesting when I went back to read it because of a few choice user comments that appeared in the follow-up comments on thier site. The thing that really struck me was not that PostgreSQL was able to out-perform MySQL, but that it seemed to do it so thoroughly. Not only was it able to run the complex queries faster, it was able to run the count queries faster, and was able to use less memory, and use less disk space, and had faster load times. And all of this even the the FeedLounge guys clearly have a fair amount of experience with MySQL (as evidenced by the performance suggestions in thier comments, and that they were all tried and then some).
It really makes me skeptical that MySQL could ever provide a capable MVCC style replacement for InnoDB should Oracle decide to put the screws to them, which I would guess would mean even more people looking into PostgreSQL in the next year or so. This might mean trouble for MySQL, if they begin to lose the enterprise level OLTP market to PostgreSQL, and people begin switching to the superior sqlite for thier simple websites (rather than use the more complex myisam system MySQL provides), what’s the market for MySQL? They have always been considered a golden example of an open source business model, and given the VC they have floating around and that they are in bed tight with some large companies, I don’t think they are going away anytime soon, but the future really seems cloudy imho.
In any case, hopefully we can cajole the FeedLounger guys to provide a few choice quotes for the PostgreSQL website, and maybe puff things up into a full case study once they have been going for 6 months or so.