Again, I'm not saying contrary, just that this post is not evidence for which group is biased. It only shows that at least one of them differs from a hypothetical baseline.

I would be interested in a study or evidence for it. The JK Rowling example is interesting, but is only definitive insofar as it shows the publisher's belief in the statement, but isn't by itself strong evidence that it is true.

I'm not sure it's a great illustration of that; I'm fairly certain that there are a higher percentage of major movie releases that have male leads (which is a problem in itself).

To make up some numbers as an example, since I haven't counted from the diagram, if 70% of major movies had male leads, and 70% of the movies in the he/him category were male leads, it would show no bias in that group for their preferred lead. In the same scenario, if 50% of movies in the she/her has female leads, it would actually show a bias, as it differs from the general ratio. And if he/him were 80 male and she/her were 40% female it'd show a somewhat similar bias on both sides.

Again, I'm not saying that any of those numbers are real, nor that your conclusion that he/him have more bias in preferred leads than she/her is false. I'm just saying that this graph and underlying "study" don't necessarily illustrate that.