With cloud storage, you have chunks of your data distributed across potentially thousands of different servers. No one server will contain the entire file (security) and the data chunks will be replicated in many locations (redundancy).
So if one (or even many) servers go down or offline, you will still be able to recover your data from the other servers.
So the absolute opposite of what you are talking about.
Not at all. Where are your apps? Where is your data? Not on the local machine, but stored remotely. It doesn’t matter if the storage is on one mainframe or distributed across thousands of servers, you still can’t do anything without a connection (and I defy anyone to assert that the internet will be any more reliable than an in-house network. And I’m not talking about data retention, I’m talking about connectivity and throughput).
There is no fundamental difference in the workstation-server model and the frontend-cloud model. The mechanics will be different, but the results are the same. The reality is that there is no reason, beyond data portability, for the cloud model (as the expense factor that spawned the workstation is no longer present... modern desktops can do just about anything you want them to do). And data portability doesn’t require a Chromebook... just a dropBox or VCS.
Only the software companies drool over the “cloud” (because it is in their benefit). And I’d like to see you open your cloud-based business in Iowa...
(In other words, you pretty much ignored every point I made in my post...)
SSDs have 500 megabit speed as of now. I don’t see that kind of speed being all that common for home connections in the next 20 years. Putting all your apps on the cloud would be slower than a traditional HD. That’s the main killer.