I would then to agree with you, except in some special case (which is actually the case I am talking about). To continue with the same example, we had last year ~200 applicants for 5 jobs. It is not the same think as having ~200 applicants for 20 jobs. For small numbers of academic jobs, one should avoid anything else than scientific excellence.

For a larger number of jobs, I still think than 50%/15% is too large. I would feel more comfortable with a softer 25%/15%. The most important point would be to identify where this 15% is coming from. To apply for a job in research, you need a PhD. There as well, 15 % of female applicants. Then one can check at the master level, again 15% (Talking about France). At the bachelor level, I don't have the numbers but I think that it is not as extreme. Much closer to the 50/50 case. The question is then: why don;t we have more female students in the master programs. And here I have no clue.

Maybe are women not that attracted by fundamental research and the amount of necessary postdocs everywhere around the world before getting a job?

To conclude, I am probably discussing a special case which is special enough for not applying general rules that would be reasonable otherwise.