The argument goes like this: traits are distributed with different frequencies throughout populations (plenty of evidence this is true), those traits may lead to differences of interest and focus in specific areas, and that leads to differences in the frequencies at which we can expect to find members of those populations represented at the top of some fields.

None of this says anything about the abilities of any individual women. There are many excellent female engineers. Nor does it say that these excellent women engineers don't face discrimination because of their gender. All it says is that you'd expect fewer women to be represented in the population of the most competent engineers (in the same way you find fewer men in veterinary fields, for example).