Oh Gawd... Read: http://web.mit.edu/linguistics/people/faculty/pesetsky/Nevin... Edit: and sure this is the Chomskians debunking the "upstart" Danial Everett but this is thorough, clear and convincing in contrast to the haphazard novelty that most Piraha discussion involves. A language with future? How...? Well, actually the answer would be just no, that's not how it works. And sure, this headline says "number" but a…
I found the discussions of anumeracy in that PDF quite problematic because they basically argued we don't know enough about the language to say one way or another. Some of the difficulties, particularly adding vs removing spoons posed significant interpretation problems for me. Of course if I was differentiate between a few spoons and a bunch of spoons, the lines would be different if adding spoons vs subtracting them. This is because the relative frame of reference is different.
It seems to me that the problem is fraught with insurmountable epistemological problems which can't be solved and therefore people get to make best guesses. In general though I would trust the one who speaks the language more fluently than the one who doesn't. This isn't to address novelty claims. Bahasa Indonesia, for example, mostly lacks relative tenses as well, and Proto-Indo-European probably did as well.