I think it's a hill that a lot of people and companies have been pushing computing tech up for many years...now we've reached the top, and we're rolling down the other side. Getting to the top required getting mobile networks "fast enough", systems small enough (because what's the point of being mobile if your system is big and heavy...might as well leave it plugged in at home or at the office), and the web (or something like it) advanced enough to be the platform on which these network apps are built.
If it were a pendulum, we'd see it swing back in 10-20 years. I can't think of any scenario in which that'll happen, even as small/personal devices get more powerful.
Note: I don't know if our destination is a good one. But, momentum and gravity (and hundred billion dollar companies) are definitely pushing us in a pretty clear direction.
Transport signals have gone from 2.5G to 10G to 40G to 100G to 200G per wave, and from 8 waves to 16 waves to 40 waves to 96 waves to 100+ waves per fiber.
The bottleneck has existed for a decade at least, and is invested in to keep pushing it out at the same time all this other stuff is happening.
Rich vs. thin client is a pendulum that seems to swing about every 10-20 years.
In the mid-'70s, a computer was a big box by IBM or DEC with a number of thin terminals attached. These had enough power to display a screenful of interactive information without having to talk to the server constantly (sort of like an early version of HTML forms).
Then PCs happened, and suddenly you could have your very own device with a slice of the power of that DEC. The client was suddenly so rich but widely available connectivity so slow -- 300-1200 bps modems -- that most people never connected to a server (except for special national networks like France's Minitel).
By the mid-'90s, thin was in again. Sun and Oracle defined a "Network Computer" spec. Apple was planning to build devices for it, but then Jobs had better ideas and the NC never really took off.
So we got ever richer clients in even more personal form factors. The iPhone became the ultimate rich client: in the App Store model, each online service provider is expected to build a native binary and deliver it through Apple. Microsoft has been plotting along similar lines with the Universal Windows platform.
At the same time Google has been trying to push the pendulum back to the thin end. With the new "Instant Apps" model [1], even Android is being turned into a web-like platform of app fragments delivered on the fly. Will be interesting to see how far they can take this -- will the Android runtime be part of Chrome eventually?
[1] http://android-developers.blogspot.com/2016/05/android-insta...