I would more say:
MySQL: semi-SQL front end to a flat file database, with some known integrity issues, written by people who are more concerned with speed than anything else. Good as a text datastore, questionable as a long term data warehouse, unsuitable as a transactional database.
Postgresql: Solid and reliable, built by database gurus for database gurus. Performance never comes before correctness, and a VERY open design process. While performance is quite acceptable for medium loads (hundreds of users, accesses per second) it is not scalable to really large apps yet due to a lack of clustering support.
Access: Prototyping tool. Should not be used in production for any data you genuinely care about.
Oracle: One of the big heavy lifters of the database world. Scales to huge loads, but will cost you a pretty penny no matter what you are using it for.
If they were lifeguards think of them this way:
MySQL: Handsome, muscled, and tan, he runs quickly into the water to save you, drags you out, then can't remember how to perform CPR and you die anyway.
Oracle: A team using several dozen boats and jetskis, grab you in a net and fly you to the hospital under a helicopter. You find out you had a sprained ankle. Bill: $2M US
Access: You fall overboard, someone tosses you a life ring. Boat never slows down, no one is sure if you made it or not, but at least you have a nice ring to hand on to. Out in the mid atlantic.
Postgresql: Waits to make sure you really are drowning. Swims out to you, brings you ashore, performs CPR, and after you wake up, spends 2 hours lecturing you on water safety.
:-)