When I first saw the Arduino I didn't see the point - after all there were boards that cost less, did more, and the Arduino IDE seemed very barebones compared to what you could do with GCC and a custom toolchain.
Then eventually I saw how much community support, ready made hardware emerged around it, to the point that after a while, not going the Arduino route was a decision you needed to justify heavily.
Same thing with the Raspberry Pi - there are commercial devices now running or more or less stock Pi hardware with some accomodations - the power of the community is just too large - you can either spend an insane amount of time getting things working on your custom SBC, or get something well-supported for free.
I hope that the same thing will happen with the Steam Machine - the pull of the community will result in a well-supported 'default' device where people (and Valve) will put in the effort to create a comparable desktop experience to the commercial OSes.
Valve already helped immensely with Wayland - it's crazy to think that the project was stared cca. 2008, and today there's still arguments to be made it's not mature yet - by investing the necessary energy to make sure games run well, the drivers are optimized, and there's a high-quality end-user library (wlroots) for writing compositors has been the push that Wayland needed.