Although I would be in favor of increased statistics on comments / votes in order to allow the community as a whole to analyze what writing styles, etc. favor different subgroups, I think having a voting system that prioritizes different voters simply due to characteristics they display isn't a healthy way to handle this issue, as IMHO it would lead to more "us vs. them" dichotomous thinking instead of viewing us as a group of individuals focused on understanding and developing rationality in our lives and others.
After all, the question is not, at least in my mind, how can we get more people onto LessWrong and not offend them when they get here, but rather to culture a intellectual pocket that fosters open discussion and self-improvement. If some people are offended when their ideas are not accepted (as long as it isn't rudely accomplished) I don't think that's something that we should address through the voting system, as rationality is at least partially dependent on not getting besotted with a single idea or concept, regardless of its value and accuracy.
Agreed. Furthermore, I think that we should consider the fact that even if there was significant selection pressure for the ability to be literate (I would argue that this is most likely not the case) that there is insufficient evidence that this pressure could result in a significant social difference so quickly.
Furthermore, the converse appears to be currently true: the more educated a women is (I would assume this relates to reading comprehension, although not necessarily so) the expected number of children she will have generally decreases (http://economix.blogs.nytimes.com/2010/06/25/educated-women-opting-for-motherhood/).