I think there was and still is potential in cloud game streaming, its more a failure in execution. In theory we have an incredible amount of mostly idle silicon in peoples homes that with raising bandwidth and decreasing latencies become increasingly wasteful. It seems to me quite obvious that eventually there will be no particular reason why we need game consoles or high-powered gaming pcs at all. We can get much better efficiency if the compute necessary to render and simulate games happens in a shared cloud system.

Not that the economy really cares about wasteful consumption now of course, but there might be natural influences that may eventually force it to care about it.

We also never really explored what cloud-native games actually mean. Like what could we do with cloud games that we couldn't on a console? What does it mean to the compute budgets, to asset streaming, to multiplayer between cloud instances, if we could more efficiently use gpu resources if multiple players use the same gpu to play the same game, can we make multiplayer games in the cloud that are impossible any other way? etc. It feels like no game developer thought about what cloud gaming actually could mean, this could be a very exciting space in a few decades from now.

I think Stadia wasn't really a product developed by remotely competent people, they did so many mistakes that it feels like almost anybody could've told them. There even were detailed early reviews of the many issues with the platform, but they never addressed any of them. Its like they burned money it was weird.

Tying it to an expensive, mediocre controller instead of "runs on anything with a modern browser" was especially confusing to me, it seemed so incredibly foolish.