I just want to point out that Scanlon has posted [scroll down] in the comments of DeLong's post conceding Weatherson's point about Delong's reply to one of my posts about utilitarianism.
The question of how “well being” should be understood is really several different questions. As a result, there is a tendency for people debating the question to talk past one another, and this tendency is to some extent represented in this thread.
One way that the question of well being might be understood is: what, at the most basic level, should an individual want his or her life to be like? Preference satisfaction is not a very plausible basic answer to this question, because in many (I would say most) cases people (rightly) prefer things because they believe them to have other properties, such as being pleasant, or worth doing for other reasons. They do not seek these things simply because they prefer them.
On the other hand, the question might be: How should the quality of individuals' lives be assessed by policy makers, as a basis for governmental decisions? Here preference satisfaction is a much more plausible answer, based on the ethical principle that governments should defer to individuals' own assessments of what is good for them.
There are lively debates about the proper answers to both of these questions. Nozick's experience machine is a contribution to the first debate: Should individuals take the quality of their experience as the sole determinant of the quality of their lives, or should they take other factors into account, such as whether our pleasure comes from true beliefs about what we are actually doing, and whether these things are in fact worth doing? Sen's work on capabilities is a contribution to the second debate: Given factors such as the adaptability of preferences, shouldn't governmental policy be judged on some basis other than facts about what people happen to prefer?
I won't argue here for my own answers to these questions. My point is just that they are different, and that this difference needs to be taken into account in order to understand what we are disagreeing about.
My view is pretty close to Scanlon's these days, so that's pretty cool.