That's for console use. Trees are useful for looking at the whole thing at once, much as it's easier to appreciate a real tree that way than by maintaining a list of the leaves and branching distances.

I'm a a researcher and my source material is ~50gb spread across ~50k hand-picked files with pretty arbitrary structure, plus another 4-5 times that in mirrors of other collections. Trees let me see things that I would overlook by reading lists all the time.

Right now I half-ass use some bioinformatics tools originally intended for cladistics and proteome research to visualize the collection. Some days I go looking for specific information or catalog what I have, other days I stumble across a cache of valuable data and might select hundreds of individual files.