Why use folders in a filesystem if you can just prefix the filename?

Well, on a file system, putting fewer files in a folder improves performance.

There's also the human factors of people liking to group things; and not have ls -l/dir scroll of the top of their screens.

But constants become hardcoded literals in the "compiled" code tree, so no gains to be had there. And the compiler doesn't care a jot how "full" a name-space is. Indeed, the whole idea of name-space "pollution" is pretty dumb: an empty name space is an entirely useless entity.

The only reason for the prefix (of either form), is to allow for multiple definitions of the main names for different purposes; and both prefixing methods give the same range of possibilities.

But one is obscure and the other not. (I'm done! It's your code :)