> What would be the options?
Soyuz uses straight hydrogen peroxide, I believe, 82% concentration. For 3 tons capsule it takes 30 kg of peroxide - monopropellant - to land.
Buran is still the example of a high tech option - it used LOX and kerosene for RCS, the biggest trick was to keep LOX liquid for a month-long missions. It was done with mid-1980 technologies.
Kliper planned to use LOX and ethanol. The craft wasn't completed in metal though.
One has to remember that hydrogen peroxide - the rocket concentrations are called HTP, high test peroxide - can be used as an oxidizer, after a catalytic decomposition - and after decomposition HTP is easily hypergolic with fuels. "Hypergolic" doesn't mean "nitrogen-containing, long-storing and toxic", it means "self-igniting" - and it could be non-toxic (or at least HTP toxicity is on the whole different level that hydrazine's one).
To come back from orbit you don't need too much delta-V. So Soyuz uses 30 kg of peroxide, maybe it would use 15 kg of hydrazine for the same effect - you're saving 15 kg in a 3 ton spacecraft but losing on toxicity. Similarly, Crew Dragon is a heavier craft, flying on a bigger F-9 rocket, it can probably allow for some extra mass spent for RCS fuel in order to make it non-toxic.
> You could go with a monoprop Hydrazine which would eliminate half the toxic stuff.
Hydrazine is, I think, a way bigger half. NTO is comparatively benign... not so carcinogenic. But still lethally dangerous :( .
> Biprop hydrocarbon/gaseous compressed oxygen?
Was done at least. Yes, it would require extra systems. You think it can't be done?
> Do you consider high-test peroxide non-toxic?
Comparing to hydrazine - yes.
> The decomposition would also make storage and rotation a consideration.
Yes. Have to chill it, take care while in storage and keep tanks extra clean. But still.
