Indeed, if you look at sex not as a matching problem between males and females, but as an event distribution, then let's say "desirable men" want sex at a rate of K events per week and all women want sex at a rate of J events per week. Let's say the population has 10 desirable men, 90 undesirable men, and 100 women. In the past where the desirable men were married off, they had sex avg(K,J) times per week, for a total of 10xavg(K,J) events involving a desirable man, leaving 90xavg(K,J) events involving an undesirable man.
Let's say now all those people move in the new, "casual" world. The desirable men can have sex to their heart's content: K times per week. The undesirables must use the remaining times that a woman in the population would like to have sex but the desirables are satiated: 100xJ-10xK. Now, if 100xJ - 10xK > 90xK, so if men and women want sex roughly as many times, or if women want more, undesirable men are not really negatively affected. Indeed they have more variety under the new system.
However if J = 0.5xK, then roughly half the undesirables are unsatisfied, or they're all unsatisfied half the time, which does feel like a step down. Anecdotally I don't think it's the case that 50% of less desirable dudes are just not having sex, but let's say it's 10, or 20%, that's believable. It would mean women want sex slightly less than men -- 10 or 20% less. Note that the 10 or 20% not having sex aren't necessarily less desirable than the other undesirables, just less lucky.
Clearly a solution for the leftover guys is obvious: They must increase all women's libido past the points where the hotties can satiate it.
(Obvious caveats: silly analysis, doesn't model gay people, assume libido is constant or normally distributed across populations, assume partners for every event are independent, etc)