Distressingly, ITER is actually on schedule per dollar with its original plan. It's the amount of dollars falling behind per unit time, rather than the science and engineering per dollar.
On the material science front, there's two issues - normal operation, and disruptions. With normal operation, there are materials that ought to be able to do a pretty good job of withstanding the environment inside the reactor, but the trick is finding ones that will do so without poisoning the plasma. Right now, there's some really cool work being done with liquid-lithium walled reactors to try and ameliorate those problems. As far as disruptions go, that's a confinement issue, there probably aren't materials that can deal with it. But almost all of the research being done with the computational plasma physicists I was working with this summer was going into understanding the magnetic reconnection events that lead to instability and disruptions. There are also reactor designs other than tokamaks which ought to be inherently more stable, but which have had tremendous difficulty getting funding due to the politicized nature of the work on ITER.
NCSX, for example would have had some very interesting results had it not been cancelled, but thankfully other stellarator experiments are under way (HSX, LHD, and the Wendelstein 7-X).
