nor that it would be good to have a general policy of smart and ethical people having less children than people not belonging to that class.
It depends on the counterfactual. If it consists of donating all the resources a kid would cost to the best cause, then that likely trumps everything. Especially also if you take the haste consideration into account; it takes forever to raise a child and good altruists are likely cheaper to create by other means. And the part about altruists losing ground in Darwinian terms can be counteracted by sperm donations (or egg donations if there is demand for that).
Having said that, it is important to emphasize that personal factors and preferences need to taken into account in the expected value calculation, since it would be bad/irrational if someone ends up unhappy and burned out after trying too hard to be a perfect utility-maximizer.
I've always been more of a theoretician, but it's important to try one's hand at practical problems from time to time. In that vein, I've decided to try three simultaneous experiments on major Less Wrong themes. I will aim to acquire something to protect, I will practice training a seed intelligence, and I will become more familiar with many consequences of evolutionary psychology.
In the spirit of efficiency I'll combine all these experiments into one:
She's never seen Star Wars or Doctor Who.
She's never seen David Attenborough or read J. L. Borges.
She's never had a philosophical debate.
She's never been skiing.
Never had sex, never been hugged or even been licked by a dog!
She has so much to look forwards to...
(Though she'll be very boring for several months yet!)