The basis for this argument is that both boys and girls are equally likely at birth to go into a given field. As they grow society gives cues to them that X is for boys and Y is for girls. These cues end up shaping preference and explain the large difference we see in the ratio of boys/girls in X. The issue I have with this theory is that it shouldn't apply only to STEM, it should work for all fields. Thus if instead of STEM we say Law or Finance or Medicine have cues then that should mess with the ratios there.
If you go back several decades the ratio of men/women in Engineering, Law, Finance, and Medicine were all equally low. As time moved on the ratios improved in Law, Finance, and Medicine but not in Engineering/STEM. If you want to claim cues as a major factor in the current STEM ratio you have to explain either
1) why unchanging cultural cues were not the reason for the improvement in Law/Finance/Medicine but still negatively affected STEM or 2) why the cultural cues when away for the other fields and not STEM
I haven't been able to find an answer to either of those points and so I think that while some cultural cues exist, it isn't the main reason for the ratio differences and doesn't need to be address to fix the issue.
The current ratio varies widely between STEM fields.
It would be helpful to understand how the culture came to accept women as lawyers and doctors, but why would that be necessary to identify cultural factors as a reason for the imbalance in CS?