I think this article demonstrates nothing other than the limits of empirical analysis, particular when it applies to human beings.
First of all, one of the reasons raising children is so frustrating is because of the society we live in. Of course, it's very expensive--and therefore frustrating--when you have to deal with things like increasingly high tuition costs for increasingly little return, which is mainly the fault of the state. The state also removes children from the parents most of the day, placing them in prison-like indoctrination camps called schools, which tends to turn them antisocial, unintellectual, and isolated from their families (and, of course, the parents were brought up in those schools to). For complex reasons, you have high divorce rates, broken families, and custody battles. You have the news media constantly making parents irrationally fearful of all and sundry harms befalling their children, driving them to obsessive paranoia. Meanwhile, social security and entitlement programs render children much less of an economic "investment" then they used to be.
So put it all together and of course it seems like having children is an irrational frustration in empirical studies. But can these studies measure emotional fulfillment? No. And the proof is in the pudding. People keep having children because they want, and they keep keeping their children, sometimes going to great lengths to do so. As the rationalistic method of praxeology tells us, they do this because having, keeping, and raising children satisfies the parents' subjective desires, which, sadly for the empiricists, can't be measured as a statistic.
And it is still a good non-economic investment. Have you ever witnessed the death of an elderly person? Who's always gathered beside their bedside? Their adult children, assuming they raised them with love and care. Who's beside the bedside of dying childless old man or woman? If their lucky, their nieces and nephews.
