There's a lot of misunderstanding here about these statistics. The purpose of releasing these numbers isn't to institute a "quota" system - it's to show that there is fairness in your hiring practice. The biggest criticism here appears to be that one can only hire the talent that is available, whatever race they may be. I agree with this - and if you're hiring practice is fair and open, the demographics of the hirees should closely match the talent pool from which you're hiring from. And for a large enough company (Apple, Google, Yahoo, etc.), the statistical deviation from that mean should be small. Incidentally, in my jurisdiction statistics like these are used to monitor hiring practices and ensure that no discrimination or hidden bias is occurring.
Apple's numbers appear to show a fair hiring practice, as their numbers at a glance match the applicant pools. For example, 10% of US college graduates are black, according to the US census survey, which closely matches their 9% of black non-tech workers. Google's and Yahoo's numbers, on the other hand, showed only 1% of non-tech workers as black. The implication from those numbers is that while the average black college graduate has an equal chance with his white counterpart of getting a job at Apple, he is 10x less likely to obtain a job at Google or Yahoo. That is where the cause for concern arises.