This article's evidence that the #TwitterMigration failed is that when he shut down his server, 25% of the servers he notified didn't respond back. I'm not sure that means what the author seems to think it means. It's not like every user has their own server. And when you let anybody build a server in a network you're going to see a lot of failed launches. Honestly, with the explosive growth Mastodon had, retaining 75% of all those new servers is pretty impressive to me.
Mastodon did not immediately replace Twitter worldwide. That's fine, growing that fast wouldn't be sustainable. But it did get a ton of new users, a lot of visibility, and enough activity to provide content for anybody who signs up. Onboarding kinda sucks, but once you follow enough people and hashtags it absolutely scratches the same itch Twitter did for me. I'd hardly call that a failure just because it didn't instantly become the next social media monolith.