Then there's the fact that heat is very difficult to get rid of when in space. The ISS's radiators are much bigger than its solar panels. If you wanted to have a very-long eva spacesuit you'd have to have radiators much bigger than your body hanging off of it. Short evas are handled by starting the eva with cold liquids in the suit and letting them heat up.
All of the mockups of starships going to Mars mostly fail to represent where they're going to put the radiators to get rid of all the excess heat.
That also makes nuclear totally infeasible- since turbines are inefficient you'd need 2.5x as many radiators to reject waste heat. Solar would be much lighter.
https://en.wikipedia.org/wiki/Spacecraft_thermal_control#Rad...
(How hot? I won't quote a number, but space nuclear reactors are generally engineered around molten metals).
The S6W reactor in the seawolf submarines run at ~300 C and produce 177 MW waste heat for 43 MWe. If the radiators are 12 kg/m^2 and reject 16x as much heat (call it 3600 W/m^2) then you can produce 875 watts of electricity per m^2 and 290 watts at the same weight as the solar panels. Water coolant at 300 C also needs to be pressurized to 2000+ PSI, which would require a much heavier radiator, and the weight of the reactor, shielding, turbines and coolant makes it very hard to believe it could ever be better than solar panels, but it isn't infeasible.
Plus, liquid metal reactors can run at ~600 C and reject 5x as much heat per unit area. They have their own problems: it would be extremely difficult to re-liquify a lead-bismuth mix if the reactor is ever shut off. I'm also not particularly convinced that radiators running at higher temperatures wouldn't be far heavier, but for a sufficiently large station it would be an obvious choice.
The Soviet ones used K (or maybe NaK eutectic); there's a ring of potassium metal dust around the Earth people track by radar (highly reflective)—a remnant from one of them exploding.