That's absolutely another major concern. I can't tell if they've designed their centrifuge to also provide structural support, such that I guess only an outer sleeve of the thing is able to rotate. Because, strangely, we're not afforded a view inside the latest build. And if I were there, in person, I'm not getting anywhere near that thing when it's running.

So they're like "Hey we can pump down really quick, it's only a rough vacuum". Well, guess what the pressure differential is between 760 Torr and 1E-7 Torr vs. 760 Torr and 1E-3 Torr? Just about 760 Torr, for both. You've got 1 atmosphere trying to implode the thing (love the music). Yeah, the thing will have to live in a web of buttressed supports outside of it.

Ideally, you want to cut a vacuum chamber from a single hunk of steel. There's not a forge in the world that has a 45 m radius capability, I just checked. ~15 m seems to be the current record. Not even close. So you'll unavoidably have welding joints in the chamber walls, and you'll have to pass through resonances with chamber vibration modes as the centrifuge spins up or down, with atmosphere pushing in from all sides. Cool, seems fine.

Ideally, you'd want the walls to be thicc af, but... how do you weld something super thick together? Genuinely asking. Surely the steel you're fusing together only extends maybe 1/4" or 1/2" into the steel, right?

Ladies and gents, I present to you - The Theranos of the space industry.