In order of importance and relevance to building a powerplant.
Materials - it is a huge goal to have materials that don't become too radioactive in use and that don't need to be replaced too often.
Cost - Big, complicated machines are expensive, presently too expensive to make money
Science - holding the plasma still is not easy. It has a tendancy to try and scamper off towards a wall and needs to be kept in place activly by varying the applied magnetic field.
In a nut shell, we could build one today that would produce more power than it consumed, but it'd be reall expensive, not last as long as we wish and not be as clean as we want.
I think it'll take another $20B before something commercial is viable. Slightly more than is spent on porn annually in the US.