This is discussed under the "argument from potential" in ethics. One problem in that argument is if potential matters when babies have it, it seems like it should also matter when other things have it. For instance, a fertilized embryo, a man and woman in a room who could start making a baby, or even a pile of organic matter that, with the help of highly-advanced future technology, could be assembled correctly into a fully-functioning adult ((let's suppose we had such technology now: would we then think piles of organic matter are similarly important as existing people?)). Those things don't seem like they carry the same value as a person.
There's a separate question on whether it's good to make new happy people and add "life years" to the world's total. Even if one thought this was the case, it seems like more of a "morality for global policy" rather than "morality for small-scale decisions." Unless you think individual humans have a duty to have children (assuming enough resources are available to make them reasonably happy), then it seems crazy inconsistent to think the added life years of a surviving baby make up for the total violation of a person's life goals. (Exception: If the adult person doesn't care about their life.) Besides, the view that it's important to bring new people into existence is controversial, even though many EAs believe it. (I wrote a post on population ethics here that discusses more of it. The way I see it, this is a view individuals can have for "What should be our global policy for the future?," but it's not some deep truth of ethics that everyone has to endorse.)
I recommend The Elephant in the Bednet as an accessible introduction to the different philosophical theories for the badness of death.