Rictor Norton has a greate deal to say about what he calls "Social Constructionalism" and "post-modern Queer Theory". The essence of his critique is the simple fact that natural appearances do not first "get real" when scholars learn to describe them in clear words.
It is of course true that borders between various categories are diffuse, but it is as true that the categories themselves very well
do exist. There are (not few) cases of intersexuals, but one hardly will have difficulties to clearly describe 99.8% of human beings as "males" or "females". Analoguesly, the existence of XX% of bisexuals does in no way suggest that homosexuals and heterosexuals are "artificial categories", as some apologets of the Queer Theory do want make us to believe.
There is nothing wrong in de-constructing prejudice and overcome gender "role models", but there is also nothing wrong in listening to one's own nature. If a boy prefers to play with cars instead of dolls, it's also all fine - no need to cure him of his "overcome masculine attitudes".