That book seemingly exclusively leans on misleading arguments. For instance, literally their very first effort is to try to 'debunk' the idea of having Mars as a sort of 'backup' to Earth by claiming that even in the case of a doomsday event Earth would still be far more hospitable than Mars. That statement is completely true but also completely irrelevant.
Take a typical doomsday event, an asteroid impact or a supervolcano. Both kill you the same way. It isn't the event itself, but rather the sun ending up getting blotted out for years by mass debris/ash not only causing an extreme freeze across the planet, but also ending photosynthesis rapidly killing all plant life which starts a mass extinction on up the food chain to animals that ate those plants then animals that ate those animals and so on.
This is the sort of event that could easily completely kill off humanity, but it's not because it'd make Earth a worse place than Mars. Even at the climax of mass extinction, Earth would still be dramatically more hospitable than Mars. The reason it will be so deadly is the same reason that more people die in the desert of drowning than of thirst. It's something you simply don't prepare for. An offworld colony in this case would help ensure humanity is perpetuated, Earth is recolonized, rescue survivors, ensure global order, and so on. In fact this is the case for most of all conceivable disasters.
If there's any argument you found particularly compelling from the book, please do share. I have a copy if you simply want to reference the page number or whatever.
Just FYI, my copy's 1500 miles away from me at the moment, so I can't give you page numbers and can only rely on my recollection of the book.
Your scenario would indeed be a "backup" for the species, but only in the very, very long term. It only works if we have a self-sustaining off-planet presence large enough to survive a complete separation from Earth. Most of the book goes into detail about what we still need to develop even to have a permanent presence in space (as in, you don't rotate people out every few months), much less a self-sustaining one.
We haven't done large-scale agriculture in space. We haven't developed methods of processing extra-terrestrial resources. We haven't seen what different gravity conditions does to children or pregnant women. We don't have solutions for the conditions people develop in zero-g, and we don't know if 1/3g or 1/6g causes those same conditions. We can't completely recycle our waste into food, water, and air without a steady supply of consumables from Earth. We don't know how to effectively deal with lunar regolith and we haven't done the engineering to keep the poisons in Martian soil out of the habitat. We haven't even developed the habitat!
The book speaks to me because I work for an engineering company and I know how much time and money it takes for even simple projects when there are lives on the line. The book doesn't say we can't colonize space, only that we have a lot of work ahead of us before we can successfully pull it off.
You're jumping from the start to the end. For instance you do not start with large scale agriculture on Mars - that will be a decades long project that will start with simple greenhouses. The first missions will be bringing 200% of the food they need with themselves, and then working to establish and ensure (as there will be unmanned deployments ahead of time) domestic life support systems. And the first habs will be exactly what we land in - the rockets themselves.
I think an important thing to consider is that in contemporary times most institutions are 100% risk averse outside of war. With Mars that will always be an impossible initial threshold. Because not only are there known unknowns, but also a practically infinite number of unknown unknowns. When we landed on the Moon internal estimates at NASA gave us about a 50% chance of success, which is why the public obituary for the astronauts was written before they'd even left Earth. Early on in the Apollo program NASA even ended up scrapping mathematical risk modeling because the numbers were always coming back so grim that they found it impossible to move forward with them.
Of course that doesn't been they were just suicidally YOLOing it. There was (and will be) extensive planning and preparation, but you have to strike a balance between achieving things and working to create an acceptable perceived risk factor. So for a practical example - the long-term effects of low g, as opposed to 0 g, can be relatively safely ignored. There's every reason to think that low g will be much closer to 'normal' gravity than 0g in terms of effects, and predictable bone/muscle deterioration can be mitigated with exercise or even weight suits much more easily than on the ISS.
---
An even more general issue is that the best place to test things with relation to Mars is Mars itself, because Mars is far closer to the Earth in just about every way than it is to the Moon, let alone the ISS. So we have things like the Mars Desert Research Station in Utah, or AMADEE-18 that was carried out in Oman. But at the end of the day it's Earth - we know the conclusion of these simulations, and there's limitations on what we can simulate.
We might have issues with bone loss in low-g environments, we might not. The point is that we don't know. These are human beings we'll be sending. Taking a chance on "it'll probably be all right" just isn't acceptable.
And that's just one single issue out of thousands that needs to be resolved before we can start having people live off-Earth. Most of it is stuff we just don't know how to do. You're trying to build a rail network with iron-age know-how. We can't even get to the moon right now, much less build a Mars outpost.
And while it's true that testing stuff on Mars would be ideal, Mars is too far away. You can't respond to emergencies. You can't quickly iterate designs for systems. You can't easily commercialize tourism. If we build bases on the moon first, we can be much better prepared to take that larger step to Mars. Sure, send a couple Apollo-style "plant a flag, take pictures, hit a golf ball, go home" missions using professional astronauts, then maybe a few missions where you have a lab on the ground and you stay for a month. But if you're sending people to stay, you need to be very sure they aren't going to die on you. Dead astronauts are heroes. Dead colonists are a tragedy, and an excuse that will be used to justify never going there again. If we can keep a permanent settlement alive on the moon, Mars will be that much easier.
Either way, there's still a lot that needs figured out first. The book outlines just a small fraction of what needs to be done. A good chunk of that will happen here on Earth. Some will happen in orbit. Some will happen on the moon. And eventually, some will happen on Mars. And it'll be fucking cool! Seriously, space stuff is fun to geek out on. Lots of interesting problems to solve. If it was easy, it'd be boring.
Hitting on the most general issue first - you could spend the next million years trying to simulate every possible contingency and issue with Mars, but at the end of the day you're still going to have many things with a high level of uncertainty, let alone the unknown unknowns. This means that going with an acceptably high risk is simply a prerequisite - period. It worked during Apollo and it will work for Mars, simply because there is no other option. So yes, "it'll probably be alright" is absolutely going to be a norm.
And the Moon is a complete hell hole that shares essentially nothing in common with Mars. You're looking at 2 week long day/night cycles that oscillate between absurd extremes of temperature of something like -300f at night to +200f during the day. And there's also no atmosphere which is why the Moon's surface looks like a teen with chronic acne. Even the smallest pebble will pound into the surface and often at quite a high level of energy. Similarly this is a big part of the reason that Moon dust is some seriously nasty stuff. Mars has a similar issue, but orders of magnitude less severe owing to the nature of where its dust came from, which is more similar to terrestrial dust sans composition.
---
As for the specific issue you mentioned - bone/muscle loss is not some unsolved mystery as the book implies. Your bones/muscles strengthen in accordance to the stresses they're under. In space there are 0 stresses so they deteriorate. The normal solution to this is weights, but in 0g that obviously doesn't work so astronauts are left doing largely ineffective and awkward elastic based exercises, which they have to spend 2+ hours a day doing. None of these are issues in low g where weights do work, and the bone/muscle loss will already be far less. These sort of arguments are like saying "Ok, we know this ship floats in 20 ft deep water, but how do we know it'll float in 50 ft deep?" Technically you don't, and you won't until you try it in 50ft of water, but ultimately there's no reason to think it won't.
Take a typical doomsday event, an asteroid impact or a supervolcano. Both kill you the same way. It isn't the event itself, but rather the sun ending up getting blotted out for years by mass debris/ash not only causing an extreme freeze across the planet, but also ending photosynthesis rapidly killing all plant life which starts a mass extinction on up the food chain to animals that ate those plants then animals that ate those animals and so on.
This is the sort of event that could easily completely kill off humanity, but it's not because it'd make Earth a worse place than Mars. Even at the climax of mass extinction, Earth would still be dramatically more hospitable than Mars. The reason it will be so deadly is the same reason that more people die in the desert of drowning than of thirst. It's something you simply don't prepare for. An offworld colony in this case would help ensure humanity is perpetuated, Earth is recolonized, rescue survivors, ensure global order, and so on. In fact this is the case for most of all conceivable disasters.
If there's any argument you found particularly compelling from the book, please do share. I have a copy if you simply want to reference the page number or whatever.