The tech is its main limitation, to be honest. The user that they were targeting doesn’t seem to require the convenience of being able to play a game anywhere–at least not at the expense of playing a version of the game with worse input delay and what often appears to be slightly worse visual fidelity. The trickle down effect of that reminds me of the same chicken-and-egg problem that VR had, minus the whole… financial gate for VR. At any rate, the tech didn’t inspire enough people to jump on board. If customers aren’t on board, investors aren’t incentivized to invest. If investors don’t want to invest, the platform doesn’t grow. If the platform doesn’t grow, customers aren’t incentivized to hop on board. Repeat.
As far as I can tell there are two ideal use cases for Stadia: people who travel a lot and who are addicted to Destiny 2, or people who want to play like two games a year and don’t think it’s worth buying the hardware to do it. Maybe they incentivize the rest of us to get into their ecosystem eventually via some sort of exclusive lineup, but based on my (limited) experience thus far, for multiplatform stuff I’d still prefer to play some single player games and all multiplayer games on a local platform, without question.
Every measurement I’ve seen has at least an extra 40ms of input delay added to a game. Some times that number is much higher. If that’s the best they can do, then the tech isn’t good enough. Certain genres will be DOA. Multiplayer feels worse across the board, even without considering how the issue is compounded by the likely fact that there’s probably nobody in your region playing games on Stadia for you to be matched with.