But if you look at animal or plant (or bacteria) species, they absolutely don't do this. They just grow to whatever the current capacity of the world is (meaning they would immediately choose the "low sustainability option you present).
This is despite the fact that most non-human species are most definitely living in a world that is being filled close to capacity.
So the argument to this solution is that it's not possible. It's not a Nash equilibrium and any set of humans that picks an option like that will simply die through natural selection, and this will not result in increased sustainability at all.
Humans are distinctive for a number of reasons, including that we've got the ability to model complex relationships and project forward into the future. Jared Diamond's Collapse looks at a number of civilizations which have fairly effectively managed to live within their means, though they do so by adopting moralities which are quite extreme, and generally unacceptable, to Western views, including infanticide and ritual suicide.
Most of those cultures are also fairly small. Playing out this game on a global scope, where communications can circle the world in under a second, people in under a day, and massive quantities of goods in a few weeks (via ship), is a rather different game.
If you're interested in more, look up David Korowicz, Joseph Tainter, or for a real blast of optimism, Olduvai Theory. Korowicz in particular looks at issues of global supply chain contagion and what that might have to offer by way of negative dynamics.
My point of "knowing limits" isn't that that knowledge be expressed consciously. But it will be established.
http://energyskeptic.com/2014/david-korowicz-2013-catastroph...
For the first, I think adopting a systems approach as suggested by Limits to Growth would be a good start, and find the writings of ecological economists (Odum, Daly, Hall) to be superior to most of the orthodxy, with Hall perhaps among the most rigorous.
In terms of the economic system itself: I don't see a solution to the present crisis. There are too many people, consuming too many resources, and yet we want to increase both values. There's a simple formula that was derived in the 1970s that describes this: I = P x A x T
Impacts is equal to population times affluence (material resource utilization per capita) time technology.
Alternatively: k <= S x A x P
Sustainability x Affluence x Population is subject to a maximal constraint k.
Which means that you can optimize at best any two of the three. Your choices are high affluence and large population, but low sustainability. High population and high sustainability but low affluence. Or high affluence and high sustainability, but low population.
Or of course, optimizing for at most one or zero aspects.
The "solutions" in this case involve deciding reducing population, accepting crushing poverty (as existed for much of human history), or wipe ourselves out.
I'm not predicting an imminent collapse of the global order, but things could turn south on relatively short notice. More likely this is a scenario which will play out over decades, possibly a couple of centuries paralleling the rise of human population following the Industrial Revolution.
Economic growth has to top out at some point, simply due to physical planetary limits. And no, we're not going to overcome those any time soon, if ever. Even planetary space travel is tremendously expensive, let alone interstellar travel. And there are multiple things we'd have to solve to address near-term challenges, with energy, phosphorus, other minerals, soil, and fresh water chief among them. Look up Leibig's Law of the Minimum -- any one constraint is sufficient to limit growth, and we're facing many.
Your past civilizations and empires weren't global in scope and yet even their falls typically precipitated long periods of darkness. And yes, life for much of human history was pretty interesting, in a Chinese curse sense of the word.
The solution in a nutshell? A man (or all humans) have got to know their limits.