It's not. The CS graduation ratio is just as bad.
If the industry were either actually or even merely perceived as systematically unfavorable to women, a natural consequence would be women being less likely to pursue education focussed on the field in preference to other fields that were less unfavorable.
One example is earning prospects, which might matter more for men than for women. Personally, I was torn between studying maths and film making, for example. I decided to go for maths because of the better money making prospects (I thought), thinking I could still go into film making later.
If you don't worry about income prospects, maybe you are more likely to choose English literature of the 16th century over engineering.
Just one example.
That claim only works if one assumes that any disparity is the result of systematic bias.
That's how I see it anyway: mainly based on the fact that women make up only around 20% of CS majors, I don't think the issue lies in the hiring practices of most tech companies.
No, but the people they are criticizing for criticizing firms hiring at industrt-average proprotions are also criticizing the industry, which is the issue.
If the industry is systematically unfavorable to women, hiring at the same percentage as the industry as a whole (which is what matching the “percentage of tech workers that are women” is) is indicative of being fully on-line with the average degree to which the industry is systematically unfavorable to women.
It would be inconsistent to criticize the industry but not firms that were dead in the middle of the pack.