Your argument is a straw man argument. While you can be an autodidact in most other fields, I will assert that CS has one of the highest misconceptions of applicability to career of anything you might study in college/university. There are, of course, a great many things people study in college and don't wind up using in their careers, but mostly it's because a) they don't expect to, or b) they just don't wind up getting a job in that field.
There is this great misconception that CS == Software Development, and it just ain't true. There are very few careers for and little need for computer scientists. There is a virtually unlimited need for talented software developers.
So while you're arguing about being an autodidact, the argument is about CS vs software development. It so happens that there is very little formal education available for software development, but there is a great amount of material available from the bookstore. Essentially any good software developer is an autodidact, but that's correlation, not causation.
So essentially the OP point stands. Men pulled books off the shelf, read them, wrote programs, and tinkered with and improved them to get good at software development, and the classroom or university environment had little to do with it. We found the books through word of mouth, reading articles, or just plain browsing at the bookstore. How many women go to the computer section of the bookstore, start pulling programming books off the shelf to browse, say to themselves, "Hmmm, this looks interesting," drop $200 or $300 to take home a stack of books, and then go home and not only read them but play with the sample code and write their own programs?
Yes, there are male developers who have never done this. They're the ones who suck, mostly, too. They take no interest in the field and never get better. And while I realize it's anecdotal, but I have met only one female developer who ever did this (out of say a dozen), compared with maybe 30% or 40% of the men.
