However I didn't go to Oxford/MIT so I really don't know jack.
All our current technology says that all future civilizations may have radically greater resources than us, but still finite. Finite resources means there will still be some sort of internal competition. Competition implies that one solution to the local competition is to pack up and head for the next star system to exploit. This logic then applies again from that system once it has been tapped out. It borders on the inconceivable that it is physically possible to colonize the stars, that intelligent life is common, and it is easy to get to "post-human-equivalent", but absolutely nobody has ever chosen to colonize the galaxy. One of those axioms is wrong (including, potentially, the finiteness of future civilization's resources).
This argument is mentioned in the paper, I'm just re-drawing your attention to it. It's a big one that is very important when talking or arguing about this; the answer must not cover 10% or even 99.9% of the cases because if intelligent life is that easy the galaxy still ends up colonized long before humans are sentient. It must cover 100% of the cases across an unknowably wide variety of intelligent life forms in an unknowably wide variety of initial conditions.
It's a tall order and most, if not all, of the self-loathing answers like "intelligent life like humans just suck so hard that they all die" really don't hold up under scrutiny very well; all life forms pollute themselves to death? Any problem we can see, some other life form somewhere will have solved. All life forms develop weapons technology and war themselves to death? Even hyperintelligent bees who early in their human-class civilization's history figured out how to share one mind such that war is impossible? All life forms ever decide to just stay home because they learn to live in harmony? Even on Earth you can see that while some humans may make that choice, other humans quite rationally never will.
And the whole "harmony with nature" thing isn't even a very rational idea if you get down to it, it is a very human form of romanticism that I seriously doubt will be a universal among the stars. Not destroying that which you need to live is rational, but posthuman technology generally implies that you no longer need your original natural environment to survive; I seriously doubt that all civilizations and individuals will fetishize their ecosystems the way some humans do, an ecosystem now distinctly more primitive and choatic than their own technology.
(One reason why it matters if a species has infinite resources is that it may mean that it is possible for a species to create it's own universe and migrate there, which would make sense for a lot of reasons, especially if they can guarantee that no other species can ever find it and muck with it. Be it a real universe or an advanced simulation, doesn't matter. "All civilizations, before advancing to the point they can colonize the stars, discover a method of creating a new sub-universe or performing infinite computation in finite time, and instead colonize that new infinite realm since it is so much safer" is an appealing answer since it solves the Fermi paradox in a way that implies a happy future possibility for humanity, but we're running out of physics that might contain such a possibility. It gets to the point where even if we figure out how to trick a string into performing infinite computation it wouldn't matter anyhow since there may be no way to manipulate a string anyhow, for example. This also falls afoul of my own argument above... every member of the species must choose this, no member should have doubts about it and decide to just safely colonize space anyhow and perhaps choose that answer in the indefinite future.)
More sustainable civilization in the universe is still better than less. (cf. Bostrom's "Astronomical Waste": http://www.nickbostrom.com/astronomical/waste.html)