For Objectivists only; I hope some day we begin to see some authors that go the rest of the distance by communicating such ideas to the philosophical lay-person.
And it _is_ a fine attempt. The problem, ultimately, is not that Kelley is anything less than competent; it is that the task is impossible.
And Kelley seems to know it. He does a very nice job of showing that the practice of benevolence is not, in fact, opposed to our real interests. But by his fifth chapter, he is reduced to telling us that "[b]enevolence is obviously concerned with our relationships with other people."
In fact benevolence is concerned with other people, period; benevolence, like justice, is an irreducibly other-regarding virtue. But Kelley's "revisionist" account of this virtue, in the end, merely reduces it to prudence. The possibility that the well-being of someone other than oneself may be _directly_ motivating to a rational agent is simply never raised.
What seems to be going on here is what so often happens with readers of Rand: the thought of even her most critical followers is drawn out of its orbit by a misguided loyalty to her ideas, and even such clearly rational propositions as "benevolence is good" have to be, if not rejected entirely, at least warped altogether out of true in order to "prove" that they're really compatible with Objectivism after all. As a result even comparatively sensible Objectivists, like Kelley, wind up offering accounts of other-regard that reduce it without remainder to self-regard -- in effect taking themselves to have shown, not that benevolence is prudent (which it is), but that benevolence _is prudence_.
The idea exercising the gravitational influence in this case is Rand's view that all moral standards are strictly "hypothetical imperatives" that follow from an a-moral and pre-moral "choice to live" -- and that therefore every moral principle proposed for any given agent must be justified _solely_ by its contributions to _that agent's own_ well-being. And Kelley et alia are at least right to recognize that something rides on this issue: give up this principle and you give up the "Objectivist ethic."
Of course, you also preserve the possibility of rights, justice, and benevolence, so the bargain isn't a bad one. But until leading Objectivist thinkers are willing to make that bargain, we shall continue to be treated to workmanlike attempts to prove that we can eat our cake and have it too, that "benevolence" isn't a "stolen concept" for egoists, that other-regard is not merely consistent with but actually _identical to_ enlightened self-regard -- that, in short, A is really non-A if only we believe sufficiently hard in fairies.