"There's no real evidence that (1) is false"
Since there is ample difference in results, you actually have to demonstrate that natural talent is evenly spread throughout the population, as opposed to waving the magic wand 'I want to believe it, so it must be true'.
Let's step away from SV and look at academia - who gets CS PhD's?
http://cra.org/uploads/documents/resources/taulbee/CS_Degree... - page 5. 86% male, 80% White or Asian (and 8% foreigners - who are probably mostly Asian). So roughly 87% for both.
There is also research (sadly paywalled) that considers whether PhD attainment can be explained by factors other than natural ability (it can't)
http://www.sciencedirect.com/science/article/pii/S0272775797...
Well, all that takes is 2 assumptions:
(1) natural talent is evenly spread throughout the population.
(2) SV is a meritocracy where advancement is based on talent.
That would naturally lead to a SV population that mirrors the population as a whole. Since it very much does not, one of those assumptions must be false. There's no real evidence that (1) is false, and so thus (2) must be false. It's pretty simple.