I'm trying to figure out how you would even begin to do an objective study of what you're looking for since by the time kids present meaningful interests, they're already irrevocably influenced by society and their parents' expectations.
And no, that's not a parsimonious explanation at all, it's one of many that fits a particular narrative. Equally plausible is that toys today are a relic of thousands of years of traditional gender role pressures that have no place in today's knowledge economy. You can trace "dolls are for girls" back to hunter-gatherer times where it made sense for boys to be given knives and bows, and girls to be given baskets and dolls. The problem is those roles are deeply ingrained in today's society even though they don't make much sense anymore since today women are more likely to build robots than men are to hunt their own food.
[1] Is a genetics study, the very thing you're asking for evidence of. That's how you would test a hypothesis that genetics is key to occupational interest -- look at two animals with identical genetics and see if they show the same interests regardless of environmental factors. If they don't, then the entire XX vs. XY distinction is a moot point. i.e. if two people with literally the same DNA don't show the same interests (as is the case for many identical twins), that's rather strong evidence that genetics doesn't mean much for occupational choice at all. And if you're discounting the study merely because it uses rats, you're going to have to throw out an awful lot of research from the last 100 years on just about everything related to biology.