What sort of existential risk is this supposed to protect against? The articles' examples are all wrong (these aren't Musk's examples):
"A billion years will give us four more orbits of the Milky Way
galaxy, any one of which could bring us into collision with
another star, or a supernova shockwave, or the incinerating beam
of a gamma ray burst. We could swing into the path of a rogue
planet, one of the billions that roam our galaxy darkly, like
cosmic wrecking balls. Planet Earth could be edging up to the end
of an unusually fortunate run."
Rogue planets are categorically not a threat to earth [0] -- the "cosmic wrecking ball" is pure fiction. And you get no risk reduction from supernovas / GRBs by spreading out across planets -- even collimated GRB "beams" are several degrees wide [1], a spot size light-years across (inner solar system is merely light-minutes across). Actually, those things aren't existential risks to a very advanced earth: their strongest effect is severe damage to the ozone layer [2], which other planets don't have in the first place..
(I'm not at all dismissing exotic existential risks. But is this a solution, or is it "We must do something. This is something. Therefore, we must do this.")
How about environmental disaster, asteroid strikes, nuclear war, and biological warfare. All of those would leave a self sustaining Mars colony unscathed even if they caused extinction on Earth. Also, at least one offworld colony is a prerequisite for the interstellar colonization that could actually protect against the threats described in your quote.
(I'm not at all dismissing exotic existential risks. But is this a solution, or is it "We must do something. This is something. Therefore, we must do this.")
[0] http://blogs.discovermagazine.com/badastronomy/2011/05/19/ar...
[1] https://en.wikipedia.org/wiki/Gamma-ray_burst#Energetics_and...
[2] http://arxiv.org/abs/0903.4710