With the widespread support for Steam/Valve on this forum because of their contributions to making Linux gaming easier, I'm now confused as to why people here are using Linux in the first place.
I personally do so out of support for FOSS software, the customizability, and actual ownership of software, which I thought were most people's primary reasons for using any Linux distro. Steam seems antithetical to all of these. The software in the first place became popular as a form of DRM, and it gets publishers to use it for the allowance of DRM on the platform. The Steam client has the absolute minimum customizability. Your account can be banned at any point and you can lose access to many of the games you have downloaded.
Whenever I game on Linux I just use folders to sort my game library and purchase any games I want to play on itch.io or GoG. On my Linux PC I stay away from clients like Steam because I want a PC that works offline, and will work if all of my accounts were banned. It's more of a backup PC.
Since Steam has every characteristic of Windows, 0 customizability, DRM, plenty of games that are spyware, I see no reason to really not use Windows instead for the much easier time I can have playing games.
Yes, I prefer many of the features of Linux distros, but using a client like Steam defeats the purpose of them. Ridiculous storage requirements due to unoptimized dependencies, having to have a background client running for some games and wasting resources on doing so.
So, why use Linux and support Steam, or use Linux and use Steam?
I work in science and at university, I was noticing that increasingly often, the kinds of computational work I was doing worked better on Linux. Often, there'd be software that would ostensibly run in Linux and Windows, but the Windows documentation and community would be pretty sparse.
The more I learned, the clearer it became that switching over properly to Linux was the way to go. It just provides better infrastructure. As an example, an area of science I feel passionately about is FAIR data principles, a list of guidelines on how we can make scientific data more Findable, Accessible, Interactible and Reusable. In practice, for me, this means I've gotten very good at using containers, which I found much easier on Linux