I'm male, 33, and I do systems software research and development. My family did not get a computer until I was a sophomore in high school. In high school, I played with technology (some webpages, lots of internet time), but I did not understand it. My senior year, I did some very basic programming in a course. But I did not start really programming until college, in a computer science curriculum. I was impressed how much more some students knew than me. But by the junior year, I don't think that edge made a difference.
So it's tempting to say that I am an example of the kind of person you and some others on the thread are talking about. But I don't think we can discount the "okayness" effect. While I did not start really programming until college, I liked technology before that, and that was not seen as unusual by anyone.
It's possible, then, that encouraging young girls to program early may help resolve the gender gap, but not because they then become young adult women who already know how to be good programmers by the time they arrive on college campuses. Rather, it could be a cultural ripple-effect where if enough young girls program, more people see it as gender-neutral, and when a young adult woman with no prior programming experience wants to take up computer science in college, no one says, "Really?"