They don't know exactly how to build a commercial reactor yet. The scale they've worked at so far tests the basic design but it can't break even (no fusion reactor has yet) or function continuously. They are going to build the minimum size that they think will achieve these goals but it will also have to be capable of testing to be able to fine tune operation. A commercial reactor wouldn't be built like that.
Also, keep in mind that cost doesn't scale linearly. Likely cost goes up with the cube of the size or higher.
That surprised me. At 70% of the size for $400M, I wonder why they didn't just try to build a commercial reactor in one go
Is this the a mid-step between a small demo and a full reactor? What do they do with the building after they've tested it?