I would in no way say that SV is even very meritocratic. Compare start-up demographics numbers to population-at-large numbers... The fact that commenters like you would read this article, hopefully think for half a second, and then still insist that SV is "very meritocratic" is exactly the problem.
The problem, I think, is external to SV. It's not that women and minorities find themselves blocked within SV - it's that they don't get into that path in the first place. The problem starts long before someone gets the opportunity to drop out of Stanford CS to pursue their dream of selling Facebook for Cats to herd-mentality VCs.
I'm not sure if you are making an intentional reference to this or not, but it does make an interesting counter-point to the assertion that mathematics is a meritocracy: http://arxiv.org/abs/1110.1556
"This is a special collection of problems that were given to select applicants during oral entrance exams to the math department of Moscow State University. These problems were designed to prevent Jews and other undesirables from getting a passing grade. Among problems that were used by the department to blackball unwanted candidate students, these problems are distinguished by having a simple solution that is difficult to find. Using problems with a simple solution protected the administration from extra complaints and appeals. This collection therefore has mathematical as well as historical value."
I don't think that's a counter point; it's a long discontinued practice of deliberately discriminating against people of a given race because other people of that race were too successful.
It's not something I support, by the way, in case the sarcasm in my above comment wasn't apparent.
It is one example (of many) of access to a mathematics education not being a meritocracy. In the linked discussions on that paper, there are people arguing that this practice is ongoing in the US (although tweaked a bit to be less obvious).
Regardless, I think it is important to realize that access to a [mathematics] education not being a meritocracy is different from mathematics itself not being a meritocracy. It is related, so this is only a partial counterpoint, but I think it is interesting to consider nevertheless. Certainly one has to keep in mind that biases in our education system will influence what we see in industry, even if the industry really is a meritocracy.
