I was talking about people (possibly) overestimating how attractive the median woman is*, though, not people failing to identify how attractive specific women are - which I think is what those graphs relate to?

okCupid lets users rate other users on a 0-5 scale from pictures; for each user, you can average together all of the ratings to determine their mean attractiveness. (They're also stored such that you can only look at women's rating of men, and men's rating of women, rather than also looking at men's rating of men.)

When you ask men to rate women on a 0-5 scale, they do it basically uniformly- about 5% of women have an average rating close to 5, and about 5% of women have an average rating close to 0, and 20% of women have an average rating of about 2.5. When you ask women to rate men on a 0-5 scale, they skew heavily towards giving men 1s. Now, for your question what actually matters is the "would bang" line, which has to come from some other source. I would be amazed if there were not sufficient men on the margin willing to bang a 2.5. According to women, the median man is about a 1- it does not seem surprising that there are insufficient women on the margin willing to bang a 1.