Staffed by who? There are actually some user facilities for materials synthesis--the problem is that if someone is making someone else's material instead of working on something that they're interested in, then I don't think that you're going to get the same level of productivity out. Also, materials synthesis takes time and running through lots of dead ends. When I was a grad. student, a postdoc from a collaboration would drive in to our lab with powders she had made--try to grow a crystal for a week (sleeping for maybe an hour or two a night on a floor in our office) and then go back to her home institution and come back in a few weeks. That just doesn't work--you don't have the responsiveness to be able to figure out what dead ends you're wandering down...Imagine if you were writing code and could only run it once every few weeks. Now, imagine that it was thousands of lines of nontrivial code and you're trying to debug it that way...