But you won't see the biggest issues anyway - atmosphere, crazy temperature gradients, potential poisons in the soil, psychology of no chance of survival if something goes wrong. Trying to simulate Mars is indeed useful, I agree, but it won't give us the problems of the actual colony.
If any single failure means people die you are doing it wrong. The way to live on mars is to dig a bug tunnel network with a lot of redundancy and have robots build everything on or under the surface. Sure, this means the only thing you notice is lower gravity, but frankly people are well adapted for earth and poorly adapted for every other place we know of.
Sci-fi loves the idea of domed city's or little domed greenhouses, but frankly Mars is to cold for this to work out. You need to waste a lot of energy to keep stuff on the surface warm and deal with a lot of nasty radiation long term.
Nobody says the first mission to Mars will be humans. We will absolutely have to deliver some machines and materials first. Which is perfect for testing the rockets.
That's why I keep thinking we could start off by building a colony on the bottom of the ocean first, the conditions there are such that it might as well be a different planet. And certain things like maintaining pressurized containers and working in them 24/7 will have a lot in common with space missions. That, or space base on the Moon.
