Bearings were a major concern in the our ultracentrifugal rocket engine. Magnetic bearings would have been nice but we stuck with ceramics for simplicity’s sake. However, Gary Hudson’s suggestion that we use magnetic confinement of the engine to minimize rotating mass within a structure robust enough to contain the equivalent of an exploding clutch is, essentially, a really powerful magnetic bearing. While it may be reasonable to discount the mass somewhat due to its dual function of explosion containment and magnetic confinement it is still somewhat concerning.
I don’t know if this Alpha Tech video is overplaying the problems with the Raptor turbine pumps or not, but it seems plausible that by pushing things to the limit in that engine, SpaceX may be heading down a Raptor flight-requalification process analogous to the Space Shuttle Main Engine SSME. If so, reducing the part count with the ultracentrifugal engine might become a lot more attractive. Possibly attractive enough to revisit that approach, including Gary’s idea.
