There might be a survivor bias here. How can we know how many similar projects didn't make it out of that cohort?
But a startup needs to make money or it can’t pay people. You don’t just need code. You need a product and you need to sell it.
There's a lot of failed start ups in general. Doesn't seem like its a feature unique to remote-only projects.
A lot of the Apache projects came from companies. Although some of these were more like one person at the company developed it and the company funded and maintained it.
Did Kubernetes start in person as in in the same room? It was built by multiple people but I don't know if they shared an office space. I thought it was that they shared a building.
All of the other ones though are pretty solid examples of team-driven OSS work.