Fusion is "cheap" if the externalities are priced in. (Of course that's currently a very big if.)
But even then, assuming a version of "fusion tech currently in R&D" realized today, we can probably build a lot of windmills and battery parks for the price of one fusion power plant.
Again, on the long term it seems ideal. (It can be made portable, abundant fuel in the universe, safer than fission, etc.)