Now it seems inconsistent with anything like our ordinary understanding of 'morally right' to say that the boundary separating the right from the wrong is to be sharply drawn infinitesimally below the very best action possible. 'Wrong' does mark a kind of discontinuity in moral evaluation, but one associated with with unacceptability. For this reason 'right', though not itself a matter of degree, covers actions that are entirely acceptable given reasonable expectations as well as those that are optimal. 'Wrong' comes into clear application only when we reach actions far enough below normal expectations to warrant real criticism or censure. (p.407)
So I've always preferred satisficing consequentialism: an action is right if it is good enough. This view has its own problems, though. Start with any action that is good enough. If we modify it, such as to increase the net benefits, the result is better and so (a fortiori) also good enough. So suppose I modify my initial action by, in addition, gratuitously murdering one person but saving two others (by giving to OXFAM, say). Surely I have not then on balance acted rightly!
But can we really say that this was all one act? On a more fine-grained individuation, we can say that my initial act and the charitable giving were both right acts (good enough), whereas the murder plainly wasn't. So there are responses available to the satisficer. But I'm not too sure how to assess them.
In any case, I'm skeptical that obligations and the like are fundamental to moral theory. At the base, there are only relations of value: better and worse states of affairs. From there we can ask about what 'practical morality' (dispositions of character, etc.) would tend to best promote the good. Moral obligation is constructed at this level: it is that minimal baseline against which individuals are properly subject to blame and social censure if they fall short. In this sense I see deontic assessments ('right' and 'wrong') as akin to rights talk. An important part of our moral practice, perhaps, but not so deep in theory.
That's a rough outline, anyway. Which parts of this picture do you think stand most in need of further attention?