I think you're placing too much emphasis on genetics. Genetics play a major role in determining the traits of an individual, but the way the individual was raised and the culture they grew up in also play a major role. The exact balance between the two is different depending on the trait, but an argument to the tune of "treat the likelihood of reproducing solely as a genetic trait" isn't going to have enough predictive power to say that people who don't want to have children are going to die off.
The same argument (falsely) applies to homosexuality. Since homosexual people don't pass on their genes, surely they must be bred out of the population? Well, estimates put the rate of homosexuality between 1% and 10%, so that didn't really turn out the way the purely-genetic argument would predict. It's just more complex than pure genetics.
The things I mentioned aren't just genetic - culture is also very heritable. So selection will work on that level, too.
The argument about homosexuality is similar to asking why amazon hunters keep getting killed by jaguars - you could "disprove" evolution by asking "why hasn't evolution just selected the ones resistant to jaguars?" But it's because getting as close to danger as possible is an effective strategy. That's one of the more reasonable genetic explanations for for why homosexuality still exists, too.
The same argument (falsely) applies to homosexuality. Since homosexual people don't pass on their genes, surely they must be bred out of the population? Well, estimates put the rate of homosexuality between 1% and 10%, so that didn't really turn out the way the purely-genetic argument would predict. It's just more complex than pure genetics.