Isn't that a power transfer problem? By that, I mean wouldn't the money spent to invent (they don't exist) these 'mini-reactors' be better spent optimizing the transfer and storage capability of solar, wind, geothermal and tidal power generation to non-optimal locations? I mean, sure, I probably want a nice chunk of fissionables at McMurdo Station, maybe someplace like Svalbard, but as a general solution it seems a poor investment.