That was my first thought, that these "computer analysts" of the 70s were glorified secretaries running punch cards through a system. It seems like I would have remembered or known a few more older female computer scientists if they were really 40%+ of the work force.
I've certainly known some. Anecdotally the ones that I knew tended to start in the mainframe world. Not so many in the Unix world. Which is kind of surprising given that I don't have many connections to people in the mainframe world.
If my small dataset isn't a coincidence, then someone who lived in the Unix world could easily be blissfully unaware of how different other environments were.