- Identifying inefficiencies in an existing technical solution
- Making people do what he wants
The things that Elon is really good at don't necessarily translate to great success in a long-term strategic vision. In a pure engineering discipline, like automotive production or aerospace engineering, they're very useful, but neither one of those things necessarily lends itself toward making a thing that is better or more appealing for the populace at large.
As an example: The Boring Company. The pure engineering problem of "how do I remove mass from the ground and create a tunnel as cheaply as possible" is a great application of Elon's strengths, and a useful problem to work on, because humanity will always need to build tunnels.
Hyperloop, though, is completely f'ing stupid, because regardless of whether an evacuated tube and car system can transport people efficiently or not, almost nobody wants to travel like this, and the entirety of the system is completely inefficient - trains are better in every case and with every metric.
In both TBC and Hyperloop, Elon wanted a thing, and convinced people to do it. It's just that there was no consideration as to whether the thing Elon wanted was useful or not.
This is Twitter. Elon looked at Twitter as a strictly engineering problem. And yes, Twitter was horribly inefficient, as evidenced by the fact that the vast majority of employees were let go and the service mostly continues to function, but my suspicion is that it is riding virtually entirely on H1B employees who are bound to the job else they are forced to leave the country. People will grind themselves to the bone for advancing human spaceflight and (to a lesser extent, maybe) improving alternative energy, but I don't know of any rational person who would do the same for Twitter, even if Elon bills it as the defacto public square.
Not to defend Elon, because I can't, but to be fair, any person reading this, if they had experienced the successes that Elon has in SpaceX and Tesla, fighting and succeeding against all of the naysayers, would have a non-trivial ego. When you combine that with the echo chamber of Elon's fans who treat him like the second coming, and who seem to honestly think that Elon (personally) can fix any problem, is it any wonder that he starts believing some of the bullshit? When he built the sub that led to him calling the rescue diver a pedoguy, and did it because people asked him to, and so many people constantly lauded him with egostroking compliments that he believed he could. And even though he didn't do anything that ended up saving those kids, he was found not guilty of defamation, so he still walked away with a win.
Fast forward 7 years and when he's forced to buy Twitter, he still feels like whatever he does is the right thing, and will do it regardless, and he will convince people under his influence to do what he wants so that he walks away with a win.
So that's a lot of text to say, yes, you're right, it is foolish. He doesn't care, he wants to win, and he'll do anything he can to walk away with a win, regardless of how it impacts anyone else, because ultimately it's not about the people who use the service, it's about Elon feeling good about himself.