Earlier quoted context omitted.
I would have felt the same way about MySpace. The reality is, alternatives can arise. Facebook won't last forever, and Twitter even less so (although one of my favorite attempted "Myspace killers" was a social network for karaoke singers)
Right, but look at history: MySpace died out (mostly), and was replaced, but by what? Facebook, another centralized, proprietary service. And then came Twitter, which was yet another centralized, proprietary service. When has an open, decentralized product or service ever supplanted a centralized, closed, proprietary one? Never, that I can think of. Linux never replaced Windows, not for the vast majority of users. It…
Some of your examples are flat out wrong: Oracle was the big dominant proprietary player when Yahoo decided to risk going with MySQL (which was great, and I can't wait for MySQL to get replaced by Postgres).
Now think how many websites are running on WordPress. It pains me to say it (because the code quality is so bad), but that's an example of successful open source project. Git is up there too. Don't like Github? Switch to Sourceforge. Don't like Sourceforge? Run your own server.
So if you define the category narrowly enough, anyone can say, "no open source project with these exact characteristics has ever had success," and be right. But because it's been limited so heavily, the example from the past will have no predictive power for the future. You need to find a different way of analyzing the problem.