However, many Oracle users don't need what Oracle provides. Of the last four Oracle projects I worked on only one actually needed stuff unique to Oracle (good replication/scaling in this particular instance). The other three could have been done in PostgreSQL equally easily. One of them could even have been done in a non-innodb MySQL since it was just being used as a data-store.
It's going to be interesting to see if Oracle can manage to keep a hold of its low end market in the years to come.