Yes there can be legitimate business cases for Oracle. There are situations where other products will be better, some where they will be equal, and others where Oracle wins despite the costs.
Firstly, the 'Shared Disk' RAC architecture. Shared disk means you can have multiple servers running the same database, either to spread workload or for redundant, and don't have any sharding to manage. Oracle handles getting the data from the disk to whichever machine needs it, and handles locking and conflicts. It has its own complications, but is pretty solid.
Secondly, it is a 'Swiss Army Knife' out of the box. Everything is included (though you may need to pay more to use some 'tools'). You've got Spatial, Document handling, Image Handling, XML. It scales from the freebie 'XE' edition, which I run on my netbook, up to the dedicated Exadata hardware. The database actually includes both a web server and an FTP server.
Thirdly, and the bit that is currently paying my income, is it has a web based programming environment built in. In Application Express (Apex for short) you develop and deploy through the browser. They are moving (slowly) into 'cloud' and what they have is a Platform-as-a-Service product that they've been developing for almost a decade. The application I work on doesn't have PHP/Perl/java/Python, but it solid Oracle down to the OS.
Fourthly, Oracle has been in the game a long time with a massive market share. Which means there's plenty of people around with Oracle skills. I worked at one place that had looked at Postgres as an option, but passed on it simply because there wasn't a solid supply of experienced staff. Also, staff experienced with Oracle have probably picked it up in a big Enterprise environment. Companies like to be able to get people with a matching background.
In many cases, Oracle is a sledgehammer to crack a nut. But it does allow a company to have a single database platform that they can use to meet pretty much any requirement they have.