I've been commenting too much on this post so I'm cutting myself off here, but if you want to continue the dialogue in DMs, feel free to message me.

I have finally gotten around to reading the paper, and it looks like I was wrong about almost every cited example of public opinion. On euthanasia and non-human/human tradeoffs bioethicists seem to have similar views to the public, and on organ donor compensation the general public seems to be considerably more aligned with the EA consensus than bioethicists. The public view on IVF wasn't discussed and I would guess I am right about this (though considering the other results, not confidently). The only example I gave that seems more or less right is treatment of minors without parental approval. This paper updates me away from my previous views, and more towards "the general public is closer to EAs than bioethicists are on most of these issues" with the caveat that mostly they seem either similar to the general public or to the left of them on most of these issues. I still agree with aspects of my broad points here, but my update is substantial enough and my examples egregious enough that I am unendorsing this comment.

I guess to elaborate a bit: The non-identity problem means that even choices that intuitively seem very morally dire when it comes to the kind of life you give your child can turn out to be morally neutral if the choice simultaneously changes the identity of the child you bring into existence. Because the results of biting the bullet on this seem so absurd to so many people, most papers in reproductive ethics kind of treat all choices about which child you bring into existence as though they are instead choices being made for the life of a single child. The reasons given vary a good deal, and there is more consensus that this is how you ought to treat these cases than why.

I think basically all bioethicists who answered this combination will say that the "loophole" you discovered counts as the same category as embryo selection morally. True it is a version of having an abortion, but it isn't the central case that the question "is it permissible to have an abortion" brings to mind, and these questions don't provide fine-grained enough possible answers to nuance your view. Again I think this style of response fails anyway, but it's difficult to produce a theory that doesn't involve cramming these different decisions into categories based on things like intention more than the specific range of outcomes without implying funny things in cases like "is it okay to not have a child at all" and "is it okay to select for limblessness in embryos".

This seems very simple to me:

If you think fetuses are not of moral concern but grown persons are, then abortion is just birth control, and embryo selection affects the full life of a moral patient (because presumably the fetus will be born and become one).

I disagree on both counts, the first because while I am pro-choice, I think fetuses are worthy of moral concern by some point in pregnancy, and on embryo selection I think the non-identity problem bites and most attempts to rescue a more restrictive reproductive ethics based on things like intention and relationship violation are not well argued. Still, I kind of think some of the explanations on this page for surprising answers are kind of some version of "bioethicists are probably cartoonishly woke in some way", rather than stepping back and thinking through the practical differences at play, and I would just caution people about that.

I'm not sure, again I haven't really spoken with my professor about this, and agree with Leah that the numbers are likely inflated. On the one hand Some ways of spelling out this position just seem to imply that yes, these deaths are as important to prevent. On the other hand, speaking less generously and more meta-philosophically for the moment, my impression is that people most likely to be comfortable with the age-neutral position in the first place also tend to be the ones willing to weave arbitrarily elaborate networks of moral cruft for themselves in order to avoid biting almost any bullet.

I'm not sure I even share your definition here, I think "disadvantaged" doesn't refer to a lack of compensation or anything else so specific, just overall whether you are below the relevant threshold of advantages. This seems very straightforward and I don't think I need a definition of disadvantage that specifically references compensation anywhere, just one that doesn't discount a level of advantage if it turns out compensation was involved in getting it. I also kind of disagree that you can just rely on "this is what words mean" anyway. I have taken very few surveys where I could just literally answer all the questions. Because of phrasing limitations, many questions are only really designed to allow uncomplicated yes/no or multiple choice answers to a few possible positions. Typically I have to imagine a slightly different version of survey questions in order to answer them at all.

Hm, something like this confusion could be boosting numbers, but I do have a professor who holds a position like this (I haven't spoken to her about it, so I don't know her exact justification). I find the position extremely implausible, but my steelman is probably something like this:

It is better to give someone twenty more years of life rather than two more years of life, but it is also better to give someone a million more dollars rather than a thousand.

We don't think, however, that it is right to give preferential treatment to saving a millionaire's life rather than the life of someone living paycheck to paycheck.

We infer from this that when we are making life or death decisions, we typically should not think in terms of deprived additional wellbeing at all, but rather the loss of something basic to autonomy/rights any being with certain minimum properties already has.

There are more details I could go into about theories that are skeptical of a deprivation account of death but this is sort of an attempted gloss of where they might be coming from, I recommend Shelley Kagan's book "Death" for anyone interested in an accessible treatment of this and other nearby issues. Again, I do not endorse this view, I think whatever commonality you find between all deaths, it is still very hard to deny that the deprivation is an additional consideration that is important enough to be decision-relevant. I just want to provide the steelman.

I think a complication is that some people answering might have a theory of justice wherein a fully just world by definition corrects/compensates any disadvantages that come with being blind. I think this view still raises concerns for people who either think that the loss of a major personal capability isn't something that is fungible with any social compensation for reasons basic to their theory of autonomy/flourishing, or people who think that justice will not demand fully compensating disadvantages like this at all. Still, I doubt 40% of respondents think the less plausible interpretation of this answer is true.