I worked in BestBuy once, their IT gave up on fixing Windows and defaulted to re-imagining disks upon first issues with Win/Office despite having a support contract.
I'm talking about ensuring some script you wrote in VB6 in 1998 still works on machines in 2019. Take that scenario times 100. It's not advertised but they do fix (or break, then fix) enterprise issues better than any entity I've seen, and definitely at that scale.
In this century, cloud apps as a managed service are sufficient for some 90 pct of office workers. Chromebook/thin clients can be wiped on the hour or chucked in the trash if they break. So, 365, which is killing it. Not an MS cheerleader but this is the right play for them.
Developers are obviously different and need a curated desktop experience.
Is there data on this?
So, so, so, so many companies have random internal desktop apps, or are dependent on external ones, that I think chucking a Chromebook at 90% of office workers is a pipe dream.
But absent data, this is all pure speculation.
A thin-client is just a desktop that can't do much of anything under its own power. There's good reason Apple keeps advancing their A-series chips, why not stop at "thin client" (running a browser), like a Chromebook? Instead, they're near desktop Intel speeds. To answer that, it's because they want their platform to actually be able to do something on its own. AR, VR, image processing, video editing, you name it. No developer is required to own it, unlike this idea that "curated desktops", which can be around the power of an Apple A13, as somehow just aimed at developers.
Agreed on O365. Microsoft can do thin clients, and has a complete stack that integrates easily to resolve more complex demands out of the box. Just because it's the 21st century doesn't mean any of these concepts listed are new, nor is what Microsoft built in the 80's, 90's and 00's valueless. Depending on requirements it's very strongly to the contrary.
The wish upon a star for free beer is strong.