They are down to 2 main problems now. The main one is (the cost of) scaling up. Fusion reactors will be more effective then bigger they are. The tiny test ones are already past break even.

The other is wall material. Apparently the radiation has an annoying ability to transmute the elements making up the wall of the reactor. They are working out a material that can maintain its bulk mechanical properties, even with random elements appearing in its internal structure.