You might be interested in Defining Women's Scientific Enterprise by Miriam Levin. She basically argues that women have been deeply involved with the sciences for a very long time, but their roles have shifted over time. She focused on Mount Holyoke, a Protestant women's college in New England, but many of her arguments are more broadly applicable.

Similar to your friend's example about women computers, women were instrumental in conducting laboratory experiments in the nineteenth century. This was viewed as women's work, because it was practical rather than theoretical. Women were also the primary scientific educators teaching science to both women and men, because teaching was seen as a feminine occupation. Meanwhile, male scientists focused on the theoretical tenets that made the women's experiments possible, and often published work that was based on lab work designed and conducted by women, who were only viewed as "assistants."

I don't want to toe the line of the 20 year rule too much, but I'm basically making the argument that Levin's analysis is more broadly relevant, and perhaps would be deepened by an answer from an economist or even a sociologist. The meaning of women's scientific labor has shifted over time, and women's contributions have usually been devalued, discounted as "assistance" or "secretarial work" rather than true scientific work. The goal posts have constantly shifted for women, and often around women, as trends change. Fields that were prominently male become female over time for a variety of reasons, and the average pay and prestige of those positions suddenly drop (teachers, archivists and librarians). Conversely, primarily female fields that have over time become majority male see increases in salary, shifting stereotypes about the complexity and prestige of the work, and the devaluing of women professionals (computer scientists).