Stuff just works on windows, I have a proxmox box with some Linux vms to run containers and I’ve tried several times over the last 20 years to move to Linux on my main pc but there are just too many faffy bits.
I really dislike what windows has become, it’s bloat ware that’s getting worse and worse, but I begrudgingly use it as I can be productive, the moment I can be as productive in Linux I’m off of windows, but even simple things like drivers are often not as good, lots of commercial software has barebones or no Linux support, there are many different package managers (on one hand great) but some have permission problems due to sandboxing when you need something like your IDE to have access to the dotnet package, also as a developer building apps/libs for Linux is a nightmare.
For example if I make an app for Windows I build a single binary, same for mac os, for Linux it’s the Wild west, varying versions of glibc various versions of gtk and that’s the simpler stuff.
Anyway I REALLY WANT to like Linux and move away from windows to it, but every time I try its hours/days of hoop jumping before I just end up going back to windows and waiting for windows to annoy me so much I try again.
(just to be clear the annoyances I have with windows are it’s constant ad/bloat ware, it’s segregation of settings and duplication of things, it constantly updating and forcing you to turn off all their nonsense AGAIN)
Not answering your comment directly, and I don’t even use Linux, BUT…
One reason a lot of us don’t use Linux even if we really want to us because it’s biggest strength is also one of its biggest weaknesses, that being it’s modularity.
There isn’t a single packaging system, window manager, file system, shell, etc etc.
This makes it hard for companies (and devs in general) to target Linux for releases. For example you want to release something for Windows, you build a single exe, apple is a dmg (I think) etc so you just build for one single platform with a consistent API.
When you want to build for Linux there can’t be just one build/package. This has actively been cited as reasons why some commercial software doesn’t support Linux, as it takes far more effort to support all major permutations of platform and package management.
So back to your question, why is Valve’s Steam OS going to help? Because it’s going to be a single platform with a single way of doing things. You can always go and replace the bits like any Linux distro but out the box it will be easy enough for vendors to support, it will hopefully also get more adoption because it has commercial support.
Look at Android as an example (I know it’s not entirely the same), but that is just a customised version of Linux, but as it’s consistent and has a single way to manage packages it’s widely adopted.
I am pretty sure Linus himself said how one of the reasons why Linux desktop doesn’t have mass adoption is because no one can agree on how things should be done, so we have hundreds of libs all doing the same thing in a different way. Valve will pick what they think is best (even if it isn’t technically the best) and through that we all have a singular point of effort and adoption to centralise on.