But why?

This was a totally unrealistic estimate, it’s probably more like trillions of R&D and launches. But even if it’s not, why on Earth would anyone spend that money when it’s so easy to and cheap to put data centers on Earth? We have tons of land! We have tons of space for making energy! And ocean! We can swap the servers out when they go obsolete in a few years! What’s the benefit to putting this same stuff in space?

It’s like discussing orbital, beamed solar power when land-based solar arrays are dirt cheap, yet barely getting funded as is. It’s an interesting thought experiment on using the higher solar flux, yes, and it abruptly ends when you start to consider practicality.