In the eyes of certain CIEIOs in this company, PostGresQL would have the same disadvantages and reputation as MySQL. i.e. it's free and running on common Dell hardware. :-/ The only reason Oracle is not being considered is because it's far too expensive. That leaves DB2 as the only one in the running.
Besides, we need to investigate clustering and PostGreSQL's clustering is an add-on. Or experimental.
So 1) buy the support package and 2), uh, whatever. :-P
PostgreSQL certainly doesn't have the same reputation. You can do a full-text query here across a quarter of a million posts in about 4 seconds. And it certainly doesn't have MySQL's optimizer and other squirrelly behaviours.
But I can't find the time to benchmark our application against our existing database, let alone convert it to another DB and benchmark it against that. Especially one that isn't on the shortlist.
And the reputation being perceived has nothing to do with performance.