There’s actually nothing stopping people from using it right now for things like S3. The way the system we implemented works is based on “adapters”. By default we use the “local” file system adapter. Other adapters are available for things like Google, DropBox and Amazon’s S3. There’s a technical process involved in adding those adapters and including the vendor’s SDK which is why you might hear about add ons implementing them. But that isn’t something to shy away from. The “add on” would just be moving files into place. It’s still very much our new file system that’s powering it all so very little can actually go wrong in terms of implementation.
For now, valid suggestion but it is best discussed in the thread I linked to above.