The "girls can't do math" stereotype is not just wrong, its false in the sense that we should just stop saying it. Its more akin to a bad told myth, brining harm where ever its being retold.

The #1 preferred profession from a woman in Sweden is economist, based on yearly polls. If one look at university education, women dominate the class room in every areas of match except one subject (abstract math), and then men only reach above 50% in abstract math after the 3rd year.

So the stereotype is not true. Its not true as in, women are as able to handle math as men, but it also not true because women actually work with math more commonly than men in real life. Its not true in any aspect what so ever, so please, please stop spreading the myth. Its only doing harm. If you need to describe a stereotype, take one that's not this one.

>stereotype is that it limits choice ("I'm a boy so I -must- play with guns instead of dolls"), which leads to inefficient use of the potential of the next generation //

Stereotypes don't themselves limit choice they enable statistical analysis of populations. It's what you choose to do with the analysis that can lead to limitation of choice.

However, if I'm tall and you're short then it's not better for us together as a [small] population to cut short my legs and give you stilts to give us a semblance of similarity - who's going to reach all the high and low places [efficiently] then?

Of course this doesn't speak to an individuals worth, but your sentence says "the next generation" and so we're looking at the population as a whole.

>To miss out on a math genius just because she happened to be a girl and was taught that she's too stupid to handle advanced math is detrimental to the advancement of the human race. //

Individuals are largely irrelevant to the advancement of the human race.

Now the question is if group A have a propensity born out of their biology to activity X but you have to ensure that equal numbers of group B are doing X, despite their propensity for that activity being statistically reduced over group A's then that is inefficient. I would warrant in a far more significant way. It's a big if of course.

No one is supporting teaching people they're stupid.

>So limit an individual's choices based on gender is as dumb as limiting them based on race. //

I don't agree that these are comparable. If I want to be a mother then my race is largely irrelevant.

However, those who wish to remove gendered play are attempting to limit choice based on gender. By not letting gender-C do activity Y _because that's a stereotypical activity thusfar for gender-C_ you are doing that very thing that you'd claim not to. Moreover you're assuming that the behaviour is bad just because of the gender of the subject.

[FWIW if you'd couched your arguments in terms of something akin to a Kantian imperative instead of in terms of benefit to the human race I think they'd be much stronger].