Heroic suicides are obviously quite different from those brought on by serious illness, grief, or an unbearable situation and in this sense are outside the scope of an investigation primarily for the purpose of preventing suicide as an undesirable psycho-social phenomenon. [p. 17, Suicide: An Incisive Look at Self-Destruction, by Jacques Choron. Charles Scribner's Sons: New York, 1972.]
Heroic suicides - or, perhaps, "altruistic suicides" - are just not the same thing as "suicide" at all.
The fact remains, however, that altruistic suicides are trading their lives for something else, so that it becomes necessary, in Velleman's terms, to examine the exchange to see if it undermines dignity. Most altruistic suicides would probably pass muster under Velleman's terms, because in many cases what is exchanged is life for life - one's life (and thereby essential dignity) may be exchanged to preserve the life (and thereby essential dignity) of another. The goods exchanged are of the same kind.
However, what about an altruistic suicide that was committed not to save a life, but for some other altruistic purpose? A suicidal act committed to save a child from rape or torture, for instance, or to prevent the release of classified information the leakage of which would result in mass suffering, cannot be said to exchange dignity for a good of a like kind. Suicide undertaken to prevent harm to another short of death must be seen as exchanging one's life and dignity for "mere" interest-dependent values (such as other people not suffering or not being raped), in conflict with the inherent interest-independent value of life. Of course, we must, in Velleman's view, allow for an exception where a suicide is committed in order to preserve someone else's rational faculties - for that purpose, unlike preventing torture, is of a kind with life and dignity (as rational faculties are the condition precedent to dignity).
Three possibilities present themselves. First, we might maintain the strange position that heroic suicide for any purpose other than the preservation of the life of others is wrong - that it is wrong to die to prevent children from being tortured and raped - but that it is not wrong to die to preserve someone's rational faculties for choosing their ends. Or, in the second case, in recognizing the moral propriety of heroic suicide, we can question whether "exchanging life for mere interest-dependent values" is necessarily a moral harm. Third, we might try to argue that acting in the interest of others in the heroic suicide case is somehow a like exchange after all.
I feel that this response will have little to say to those who see no problem with the first option, and can maintain a position that appears so strongly counter-intuitive and contrived. The more interesting question, for me, is whether an argument can be made that sacrificing one's life in the mere interests of others - unconnected to maintaining their dignity - is somehow different from sacrificing one's life in one's own mere interests.
There seem to be cases where sacrificing oneself in another's interest would be horrible, perhaps even so horrible as to cheapen the value of human life - such as dying to prevent minor property damage. There cannot be a blanket exception for suicide for the benefit of others. What the distinction seems to me to be is the strength of the interest - dying to prevent or relieve great suffering, in oneself or others, seems to be a morally acceptable option, whereas it's easy to see how dying to prevent someone from chipping a nail could be morally objectionable.
Velleman indicates that suicide is wrong, even to end severe pain, as long as the pain isn't so severe as to interfere with one's rational faculties. I would like to know if it is also wrong, in his view, to die to end severe pain, or prevent serious suffering, in others.