But interest is an outcome, influenced for example by the presence or absence of role models or the impression of certain fields as not welcoming to women.

If I remember the state of things correctly, the gender imbalance of interest in science is not present at early ages but develops around junior high ages when kids start picking up on which jobs or fields are "appropriate" for males vs females.

I don't think anyone is saying the gender imbalance is purely a result of hiring discrimination (although it likely plays a role), but there are clearly other biases in force that directly arise from the gender imbalance. How do you propose getting rid of those biases except by trying to "prime the pump"?