Speaking personally, I don't negatively weigh non-aversive sensory experiences. That is to say, the billions of years of unsatisfied preferences are only important for that small subset of humans for whom knowing about the losses causes suffering.

If I understand you correctly, the problem with doing this with negative utilitarianism is that it suggests we should painlessly kill everyone ASAP. The advantage of negative preference utilitarianism is that it avoids this because people have a preference to keep on living that killing would thwart.

It's worth pointing out that negative utilitarianism is incoherent.

Why? For the reason I pointed out, or for a different one? I'm not a negative utilitarian personally, but I think a few aspects of it have promise and would like to see them sorted out.