Bernard Williams on Professor George's Hypothetical

In thinking about the very good questions that Professor George raises below, I was reminded of some of Bernard Williams's arguments in his critique of utilitarianism (the "against" of his classic work with J.J.C. Smart).  Readers are doubtless familiar with Williams's points, but briefly to refresh the memory, one of Williams's examples deals with Jim, who is given the choice by a strongman of either shooting one person, or else shooting no one, in which case Pedro will kill twenty people. 

Here is Williams:

"[W]hat occurs if Jim refrains from action is not solely twenty Indians dead, but Pedro's killing twenty Indians . . . . On the utilitarian view, the undesirable projects of other people as much determine, in this negative way, one's decisions as the desriable ones do positively: if those people were not there, or had different projects, the causal nexus would be different, and it is the actual state of the causal nexus which determines the decision . . . . The decision so determined is, for utilitarianism, the right decision.  But what if it conflicts with some project of mine?  This, the utilitarian will say, has already been dealt with: the satisfaction to you of fulfilling your project, and any satisfactions to others of your so doing, have already been through the calculating device and have been found inadequate.  Now in the case of many sorts of projects, that is a perfectly reasonable sort of answer.  But in the case of projects of the sort I have called 'commitments,' those with which one is more deeply and extensively involved and identified, this cannot just by itself be an adequate answer, and there may be no adequate answer at all . . . . It is absurd to demand of [the "committed" person], when the sums come in from the utility network which the projects of others have in part determined, that he should just step aside from his own project . . . . It is to make him into a channel between the input of everyone's projects, including his own, and an output of optimific decision; but this is to neglect the extent to which his actions and his decisions have to be seen as the actions and decisions which flow from the projects and attitudes with which he is most closely identified.  It is thus, in the most literal sense, an attack on his integrity."