Depends. There obviously has to be a large change in the demand curve at the supply point if that point is fixed. But I agree that a small horizontal shift in a demand curve can have a big impact.

I don't think that's the right way of looking at it, though, because it treats the men and women supply-demand problems as independent. As long as the prefer-men employers are outnumbered by men, they won't affect equilibrium in this simplistic model.