Sure, research costs a bit more, but not necessarily that much.
Research costs a lot more when you need to make multiple trials and prototypes to build one thing. ie multiple prototypes to test multiple variables, scrapping each prototype. That is a fine way to research building a setelite or whatever.
In this case though, they're building multiple buildings, so can design each one differently, with different materials, etc (ie each embodying different variables) so that they'd end up with a living experiment. There is no need to build one, test for a year, then demolish it and try another. Sure, that would cost more than cookie cutting the buildings off the same design, but not by factors of 10.
wtf is NASA doing this sort of research for? It is outside their brief. Surely there are other agencies which can do this experimentation more effectively (ie. more research per $).
I can smell the pork and mythical $16k toilet seats from here.