I'm pretty tired of the Dropbox line. it's not like some sort of logical fallacy to say that something that can be built with existing products isn't viable as a huge company. Dropbox smoothed a bunch of edges on a VERY thin interface: files. This is creating a very thick interface out of thin air for something that already exists and works. Also Dropbox wasn't the 4000th ding dong to try to "fix" email.
They were all legitimate concerns, though. Making Dropbox "work" required overcoming all of those issues, and they did. Maybe Relaybeam can make it, but he'll need to have answers for those issues, and much more to do so.
EDIT: In fact, they have far more to overcome: Dropbox didn't have to overcome negative network effects. E.g. people will ask why they can't email instead, whereas giving people a url to get at a file did not make things worse.
When Dropbox was new, I tried it for a bit, but the only real use it got was for free website hosting. Now that I know how corporations fuck us all over and have $5/month to spend on a VPS, I would definitely share files using something self-hosted instead. I think for my use cases, any web-server works fine, for the peace of mind of not getting fucked over unpredictably in the future (which Dropbox did, when they stopped doing free website hosting)