In case you've not checked out Slashdot in the last day or two:
http://developers.slashdot.org/article.pl?sid=03/05/23/1826202&mode=thread&tid=137&tid=198
I tend to agree with the viewpoint that getting anything useful out of
such a merger will be a difficult project.
Anyone suppose they both felt they had to do this because they were each
losing ground to Postgres? If they do make a successful merger of the
best parts of each DB, they'll certainly become a much better combined
competitor ...
regards, tom lane