We may always face existential risks. We now know of some concerning the next 50-100 years, but in 100 years we may know of different ones. Planetary redundancy conceivably aids our chances with those risks as well, so is a project best begun now, whatever the lead time. I don't want us always just solving the present crisis. We should think about the distant future too.
If you ask me, focusing on colonizing planets is the completely wrong long term survival strategy anyway. What would be much better would be finding a way to survive in space, just orbiting a star in some kind of craft. I mean, if you can't grow food on it anyway than a planet is just a really big asteroid with an unreasonably large gravity well to deal with. If and when we eventually work out a way to viably travel interstellar distances (my money is on ditching biological bodies), every star would be a viable colonization target.
There is the possibility of growing crops in Martian soil. [1] Nevertheless many agree with your conclusion that stations in deep space offer decisive advantages in their low-energy access to asteroid resources. [2]
Regardless of which path we take, the immediate technological prerequisite is the same: an economical way to send Earth payloads to space, i.e. a fully and rapidly reusable heavy-lift launch vehicle. That's what SpaceX is developing.
Maybe you're right, but I'm thinking more along the lines of "autonomous life outside of earth is impossible in the foreseeable future, so optimizing for existential crises that threaten life on earth is pointless". The only thing we can avoid is that we ourselves destroy earth. Basically, if an asteroid hits us or something it's over anyway.
I think you may be overestimating the lead time to establish a self-sustaining colony. With the will, resources, expertise, and some luck, I think it could theoretically be done inside of 40 years. More likely it'll take quite a bit longer. But we don't need to be talking centuries.
