Africa, said the other day on another list: a little tabula rasa will just sit there unable to learn anything. If you don't accept a specialised language-acquisition device, you'll have to credit humans with a universal general-purpose learning mechanism, but that doesn't account fully for the observed common features of unrelated languages.
I perfectly agree with you that language is innate (whether there is an innate relationship between the language faculty and the hearing / speaking ability is another issue) but this does not mean we have a universal phonological structure in our head. I honestly don't know what we have in our head, but I am convinced that most if not all of phonological 'universals' can nicely be explained on a purely phonetic ground. What is innate is the perfection of our ability to adapt ourselves to our phonatory organs.
How would _you_ account for the total absence of clicks from languages outside the African "Clickland"?
That is an interesting issue -- I would attribute this to coincidence. Bantu languages in contact with Khoi-san quickly acquired those clicks. If history had been different... maybe the entire world would click !
The issue is -- why aren't clicks a part of the basic segmental inventory of all languages if they are perceptually so easy to distinguish ? Pronouncing a click on its own is easy, but to accompany it with a vowel is more difficult articulatorily ; using clicks as consonnants is more difficult than using them as simple sounds in communication.
I have an idle proposal for the clicks in Khoi-san languages (that is not testable but still nice to think about). Language is more ancient than the homo sapiens (maybe even more than the homo genus), it was independently invented progressively by different species of hominids. The modern human languages are polygenetic and are descendent from languages invented by different species of men. One of those proto-languages used clicks and handed it down to now because clicks are, as you observed, extremely resistant to sound change. The others did not. They didn't developp those clicks either because, well, they did not think about it. Clicks already had a use as a sound (like whistling). Their languages were already developped, so they couln't have introduced an entire series of new features in them. Introducing clicks in your language is not possible without an explicit reason like borrowing. Click cannot come from phonetic changes in normal consonnants.