I don't like the "right tool for the job" analogy. When you use a hammer to put in a nail, the hammer isn't a part of the building when you're done with it. But when you choose a DBMS, you're stuck with it in your architecture for a long time and it's a huge cost to migrate.
(I'm a postgres community member, so of course I'm glad it's working out for them, but I just don't like the analogy.)
I think its less hammer more nail vs screw. Both can hold wood to a wall and will be part of the building when you are done. For example a nail can splinter the wood which can cause a break down the line while a screw wouldn't encounter such an issue.
That goes even further, a nail gun can be much faster than a drill when doing large construction, but is more dangerous and easier to hurt yourself or mess up with.
Ie. MySQL is easier to get started with (well, was, nowadays it's much of a muchness), and is easier to scale in some situations but you trade for some other features and useful things.
Screws and nails are both useful, but I wouldn't just rely on one for every project :)
(I'm a postgres community member, so of course I'm glad it's working out for them, but I just don't like the analogy.)