Well, the first question to ask would be 'Have all human cultures for the past 6 millenia had the same gender roles?' If not, then the idea that one set of gender roles has worked well for everyone up till now would be obviously incorrect.
And a brief investigation will find that gender roles have varied throughout history. Some cultures have allowed women to hold political power, others have forbidden them any public role. Roman women could own property independently, and some became quite wealthy businesswomen; while women in Athens a few centuries earlier were forbidden to engage in financial transactions bigger than those needed to feed her family for the immediate future. In Europe a couple of hundred years ago, children would have remained the resopnsibility of the father in the case of a couple splitting up; in pre-European Iroquois society, they'd stay with the woman. Clearly, gender roles and relations have not been fixed historically.
The second question to ask is, assuming there has been a dominant pattern of gender roles throught history, whether this is any reason to assume it's the best way to continue. Modern science is a very recent invention, and a departure from much of history. It would seem silly to say it would have been better to rely on superstition instead of science, as this worked so well for much of human history. Science, as a method of finding out about our world, has worked better, and produced the practical benefits to show it. Equally, slavery has been commonplace throughout much of human history, but no one would consider this an argument for maintaining it as a legal institution today.
If you want to argue that a particular form of gender roles would be better for society, you need to explain why. Whether or not it's been common in the past is irrelevant.