I'd assign a low probability to this hypothesis. Most of the Flynn effect seems to occur on the lower end of the IQ spectrum moving upwards. Source. This is highly consistent with education, nutrition and diseases hypotheses, but it is difficult to see how to reconcile this with a sexual selection hypothesis.

Also, I'm not sure that your hypothesis fits with expected forms of infidelity. One commonly expected form of common infidelity would be generally with strong males while trying to get a resource rich males to think the children are there's If such infidelity is a common pattern, then one shouldn't expect much selection pressure for intelligence, if anything the opposite.

The fraction of the population which engages in infidelity even in urban environments is not that high. Infidelity rates in both genders are around 5-15%, but only about 3% of offspring have parentage that reflects infidelity. Source, so the selection impact can't be that large.

One thing worth noting though is that one of the pieces of evidence for disease mattering is that there's a correlation between high parasite load and lower average IQ, but your hypothesis would also cause one to expect such a correlation since reduced parasite load would be better correlated with better medicine and more functional urban environments in general. This is evidence in favor of your hypothesis.

I'm not aware of any obvious way to test your hypothesis. I'd be curious if you have any suggested things to look at or if anyone else has any ideas.