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.