Damnit, I'm up to 3.5k words in my own moonbase proposal/summary and probably only two thirds done (including editing). It's written as an OP but I guess I'll post it here when it's finished.
I've chimed in on fusion here before, so I'll just focus on that for now.
There's just the little problem that deuterium tends to react with itself to create tritium when superheated, so that you might get deuterium- tritium reactions in a deuterium - He3 reactor. (This means that you need a different design than the current tomahawk one). Problem is that there's very little He3 around on earth, so most research focusses on Deuterium- tritium fusion.( I think there's only 1 working Deuterium- He3 reactor in the entire world)
1) It's
tokamak, a transliteration of a Russian 'word', which is in reality a physicists acronym (eg, doesn't use first letters of all words and doesn't even have a single definite meaning).
2) D-T reactions are focused on because they are the most possible. They have the lowest ignition energy so we can actually achieve ignition on reasonably sized machines.
When I say reasonably sized, I'm talking about many-tonne chunks of steel using magnetic fields large enough to generate currents powerful enough that when the current is dumped into the reactor mass (which happens and isn't good) the entire mass jumps into the air. They literally have gaps of several inches into the construction so that it doesn't deform and destroy itself. And ITER, the new generation, is going to need to be far bigger and more powerful than that to achieve real ignition. That is using the best technology we have in almost every facet of the machine.
A He3 reactor would need to be even more powerful to manage power output in that manner. The technology just isn't there yet.
3) A He3 reactor would produce around the same amount of 'heat' in hot alpha particles as D-T fusion, used in both models to keep the plasma hot and burning. It doesn't need the neutrino capture and generation layer, which is how a D-T plant would create electricity, but it would need a way to take high energy protons to be converted directly into electricity. An easier problem but not entirely negligible.
Some other points;
The radiation risk isn't from Tritium. It is radioactive and so needs control and containment, but a fusion reactor uses so little fuel (measured in grammes) it's more like medical isotopes than fission fuel or waste. The biggest problem is neutrino output.
Most of the energy from D-T fusion is in the form of fast neutrinos. These are both hard to capture (no charge) and have a tendency to activate other materials. Finding steel compounds that can stand up to high neutrino flux without becoming a) brittle and b) radioactive themselves is a very hard problem, and one of the biggest remaining in the way of creating a fusion reactor.