You can't really just extrapolate the current cost of launching materials by the kilogram. A project of that magnitude would generate economies of scale and probably some innovations along the way that would bring down the cost. The "nearly perfect" fusion engine is really the hard part.