A long post I know, but comprehensive I think.
I take your subjective emotive position as primitive and undeveloped and rife with problems because it doesn't offer a reason (as it's emotive) for me to accept your position. If you like murder and ice cream, but I don't, I don't know how you're going to convince me of either. We're just dealing with preference under your theory.
I'm not denying an immediate intuitive reaction people have when faced with moral issues, like feeling repulsed by murder. This is not an entirely rational reaction I'll admit, but it's not entirely emotive. There are good reasons, after all, for believing murder wrong, as in it would destroy society. Matters of conscience are more complicated than just emotive preference for things, like ice cream.
I called your position primitive because I do agree that we start with these intuitive reactions to situations, but we then derive principles for deciphering the morality of hard cases. Utilitarianism and Kantianism are two efforts of providing such principles. I think we all agree that few if any actually keep the categorical imperative in their head at all times and use it to decide right from wrong, but that's not to say it might not describe the process many undertake intuitively.
We also have to admit that some often feel emotional repulsion to things that they morally ought not feel such repulsion for (e.g. homosexuality, mixed race marriages) and we must admit that some feel a lack of emotional repulsion when they morally ought to (e.g. child molesters, serial killers). The idea that we can logically convince the morally misguided to change their emotional preference makes as much sense as logically convincing someone to like ice cream who doesn't. We do, though, change people's minds when it comes to moral questions, which means something more is at play than simply emotional reaction.
In the examples I gave of people having an inappropriate moral compass, all have a certain underlying principle that is being violated. Namely, each shows a lack of respect for autonomy and deprives people of the power of their own decision making. This principle that drives much of moral theory must therefore be applied consistently throughout other moral decisions. So, for example, if I find homosexuality abhorrent, my mind could be changed by pointing out that my moral rejection requires that I ignore the moral principle of affording people the same autonomy I insist upon providing people in all other situations. Assuming I'm reasonable, I then will reconsider and then take a permissive view on homosexuality, perhaps while even maintaining my emotional repulsion to it. It is the logic, not the emotion, then that drives the final decision.
So, back to abortion. If we accept that we must protect individual autonomy at a certain level in order to be moral people, we then must figure out who has the right to this protection. We generally say that people do, and for reason, we must decide who is a person. The fetus is a hard case because it tests our ability to offer a fine tuned definition, but find a definition we must. Throwing our arms up (ala
@Banno) to the notion of definitions is too easy. We all know the limitations of definitions and we all know the problems of essentialism, but just because we can't figure out an exact and always accurate definition of a cup doesn't suggest we don't know when we have a cup and when we do. My response then is as it was, which is that we have to offer a definition of "person" that liberally protects things that might not entirely be people, simply because the destruction of something that might be a person is so morally wrong.