No, the risk v reward curve has to be better than an equivalent risk v reward curve for a terrestrial species survival project to justify the mission.

I would argue that just about any PhD in epidemiology, disaster relief, or geopolitics will have a risk/reward curve far better than any Mars mission.

People keep making this argument that, absent any other information, a two planet species is more robust than a one planet species. But that's not the choice we're facing. The choice is a two-planet species versus a more prepared one-planet species.

That's not the choice. Those two options are not mutually exclusive.

I think it may take some time - centuries even - for colonizing Mars to "pay off" in terms of reduced extinction risk. But I think in the long term - say 500 years from now - we will indisputably be safer as a species if we started colonizing Mars 500 years ago than if we didn't. And if we don't start now, then when?

I would increase that number if someone could tell a story with even a whiff of truth about how it could happen.