Humans with our great variations of behaviors has both pair bonding behavior and non-pair bonding behavior, with genes that code for both. If one strategy was always superior to the other then we would have settled for it a long time ago.
(Its a law or something when talking about genes that the behavior is always a gene-enivorment interaction, and there are likely many more factors that influence pair bonding behavior).
I remembered a case of of a obstetrician impregnating multiple patients but a Google search(1) easily proved this was not an isolated case as the first few links are all separate incidents.
(1) https://www.google.com/search?q=obstetrician+used+own+sperm
As immoral as this strategy sounds I admit it is probably the most effective strategy from an investment/returns pov. It is cheaper than becoming Genghis Khan(2) and the effect is comparable.
(2) https://www.nature.com/news/genghis-khan-s-genetic-legacy-ha...
"R-selectors" have many offspring, investing fairly litle in each one. K-selectors have a few offspring, but invest a lot of time/resources in caring for each. There's a bit more to it than that--and the theory has needed some adjustment, but...
Game theory should always be asked when biology is involved. I am not sure myself on the answer, but it seems more plausible than the article's "teenage girls are annoying" theory or the speculated "son preference" theory. If there is an optimal strategy, such a strategy may result in a small biological effect which makes parents look for a new chance of reproduction once the majority of child investment is done, resulting in the statistical data that shows 20.12% becoming 20.48%.