There is a lot of meat in Brian Robinson's message, and I certainly am not qualified to comment deeply upon it.
One small twist that may help is to remember that its not "math" as we today think of it that may have been selected for, but much more broadly based reasoning skills that allowed cultures to develop math in its modern sense. To the extent this is true, then human groups who have not been exposed to modern math may have no more trouble learning it (at an early age) than anyone else. So the example is a nice one in that it reminds us that present human skills (e.g. maths) and genes are linked by history and culture.
Brian Robinson and Davies each have interesting things to say about "emergence" and "self-organization" across increasing levels of complexity. I suspect that we are just beginning to deal with these issues squarely. Indeed, we are probably still way off target, but the effort is worth it. And its fascinating.
On the issue of alien intelligence, etc., I suspect that we should keep our minds open to the possibility that combinations of basics different from those we (humans) possess can, in principle, lead to forms of "intelligence", etc. We might not be intelligent enough to recognize these, however!
Much of the work in complex systems theory seems to be trying to get at the idea that different constellations of "elements" can produce similar products (as in evolutionarily analogous characters), just as similar initial constellations of "elements" can diverge (as in evolutionary "homology"). Given that, it is a challenge to know what to look for, on any time scale, or at any level!
But the search can be rewarding, as long as one has a certain tolerance for ambiguity, not to mention blind alleys.
John