I'm going to give a tentative answer full of speculation and guesswork, but it's too long to fit in a comment so here goes.
Sex duration is possibly a sort of human reproductive handicap. Sexual arousal can soften the cervix, increasing fertility during lengthy sex sessions. Lengthy sex exposes them couple to predators and takes time and energy, so if you have the calories to burn you are probably a better mate.
The more orgasms a male has had in the recent past the longer the average sex session takes to ejaculation. This kind of makes sense as an adaption strategy to different frequencies of sexual intercourse (if less sex happens it's more important to ejaculate quickly, if more sex happens you want to get as much fertility out of the sperm you produce as possible).
Comparing testicle size and penis size to other primates, humans have a large penis and relatively small testicles(compared to chimps, say) which implies we spend more energy on the sex and less energy on the ejaculation than chimps. It also implies a lower sexual frequency or smaller ejaculate sizes or larger vaginal depths or really a number of other things.
The whole system kind of makes sense. Long sex durations are a sexual handicap system to show the fitness of partners, which are rewarded with slightly higher fertility. The whole post-sex evolutionary mechanism thing is a layer on top of the underlying mate selection which is beyond the scope of a book, nevermind a stack exchange answer.
That's probably at least a little bit wrong. The handicap principle isn't super well accepted as a general evolutionary principle, even though it explains some things.