Actually, you can, they did, and any developers who thought they were entitled to anything from Twitter don't understand platform risk and learned a lesson.
That's like saying, "There's virtually nothing keeping people on Earth if we develop cheap, reliable, faster-than-light space travel."
All I'm talking about is gradual migrations like the ones from geocities and myspace. In the case of those services users had a lot of content they'd created that they had to leave behind. By it's ephemeral nature Twitter has very little of that.
You can build a rival service if you want, but nobody will use it until it provides value, it won't provide value until people use it.
Also, MySpace was a technical fiasco. (Although it wasn't quite as bad as Friendster, which was so slow it was unusable.) Facebook and Twitter are never going to allow pages to become painfully slow, have unreadable backgrounds, or blare music automatically.
Thirdly, Facebook's network effects were more powerful than MySpace's. Their school-based distribution strategy guaranteed that your relevant friends would be on the service. Features like "events" gave everyone a reason to join. And, ultimately, Facebook appeals to everyone age 4 to 100+. MySpace's main appeal was to music-loving teenagers. The more people you appeal to, the stronger the network effects.
Long-story short, it's not impossible to beat Twitter at its own game. But it's really, really fucking hard. It's not simply a matter of building a slightly better Twitter. Likely, Twitter would have to fuck up their basic user experience in a major way, AND you would have to devise an extremely superior user acquisition model.
Just like the developers would ditch Twitter for the next big thing if it ever pulled a MySpace.
There was a mutually beneficial relationship that is no longer (as) beneficial to one of the parties.