I don't think that the development of that kind of hardware is the worst part. Also it seems a good idea to set priorities : first the launcher, then the spacecraft itself, then the modules. Else you end spending money on a lot of fine ground-tested hardware that will at best end in a museum. Its also why it takes so much time : they can't develop the whole program in parallel just as it was the case for Apollo, as they simply don't have the money. And, who knows, if the geopolitical situation was to change in the 10 next years, international cooperation could still play a role : Russia has a long experience of habitation modules.