We are mostly stuck in eternal bikeshedding.
That feels to me like a lot more than bikeshedding (assuming there's an answer other than "zero point zero zero")
My current view on this is that it makes more sense to see it as a distribution of personality types and that gender is a red herring. It takes a fairly specific type of person to pursue computer science and coding for intrinsic reasons, and it is difficult enough that it takes some degree of intrinsic motivation to do well enough at it for it to be profitable. As it stands, the salary draw is already as high as it probably can be (in fact, I believe salaries to be in a bubble) in the US at least, and there are enough scholarships, programs, and free resources that almost anyone who is even somewhat motivated and not socio-economically crushed at the onset can start the process with some degree of ease and the expectation of high rewards at the end.
Hence the extrinsic draw of the profession is already as strong as you could expect. There is still cultural stigma, but that is being erased naturally since programmers have a high socio-economic status for the most part and indoor activities are more and more popular and mainstream. The only real problem left is intrinsic motivation, and if it is tied to your personality type as per my hypothesis, I don't see how you could meaningfully change it. Or to be more precise, the mechanisms of change that are available to us are already well underway.
> there are enough scholarships, programs, and free resources that almost anyone who is even somewhat motivated
This might be the hinge upon which we could pivot though. Does everyone really have the at-home internet access, quiet location to focus for 60-90 minutes at a time, with a desk, screen, and keyboard such that adding a Raspberry Pi is enough to get started? I can literally reach out and touch my 2 27" monitors, my work laptop, my personal computer, a KVM switch, and a Mac for each of my primary-school age kids, each with their own 27" 4K monitor, keyboard, chair, etc. That, coupled with my career as a model of what "normal" is, makes it a lot more likely that they'll end up in computing than the average Cambridge, MA schoolkid.
I don't think anything of what I described is "intrinsic" to them. If that's the case, there's chance for intervention to beneficially change things (rather than simply over-riding intrinsic preferences, which I think is bad.)
> and not socio-economically crushed at the onset
This is another one, of course, partially touched on above, and definitely inter-related.
To be perfectly clear, I am not making a by-the-bootstraps argument, just trying to look at the situation as objectively as my biases allow me to. It might be far, far easier for the scion of a tidy bourgeois home to make it, but unless they have the right personality type they are still highly likely to fail eventually. They'll go further than a student with the same personality type but fewer resources, but the roadblock will still appear later in the sequence, possibly at the very end. Conversely, a disadvantaged kid with the right personality type will typically have the possibility to get to the launchpad zone towards a better life. At the very least, the conditions and probability distributions have never been better compared to other careers and aspirations.
What could change this is to have better opportunities for mediocre coders, but it appears to be going the opposite way. Due to the extrinsic rewards there is now a glut of such employees that might pop their own code-monkey bubble soon while leaving the outrageous tech salaries for the higher tranches intact for at least a bit longer (if there is in fact a separate bubble as well for the latter).