If it's substantially more expensive per watt than solar, wind, and so on but can still be done compactly it could make sense for deep space probes that are too

If it's substantially more expensive per watt than solar, wind, and so on but can still be done compactly it could make sense for deep space probes that are too far from the sun for solar power. Currently these are powered by capturing heat from decaying radioactive elements which is far from efficient or cheap. Mostly I think that's just nasa justifying nasa doing this research though.