The real lesson from Dropbox is that things that seem technically easy to implement are in fact bedeviled by all sorts of hidden little gremlins, and finding and killing those gremlins is valuable.

Lots of services can sync files and folders. Having used many alternatives (One Drive, Google Drive, iCloud Sync, even scripted rsync), it turns out I still pay for Dropbox. It’s the most reliable for my use cases, and is not too expensive.

There is a huge difference between “I can conceive of how to implement this feature” and “this feature works for years on end at large scale with adequate reliability and usability and low enough cost.”

This is a different point. We’re already talking about a company trying to build the same feature set, not an individual doing it themselves using existing technology.

Storing and moving TBs of data doesn't sound that easy.

I'd rather pick insert an image in a document and keep a log of it as a weekend project.