I recently visited Seth Godin's Web log to dig out his item about yak shaving, and while I was there I saw this recent posting about being missed. He asks the question: if you didn't show up, if you suddenly went away, who would miss you? And he proposes that it might be a valuable goal to make it so that a lot of people would miss you. That's certainly an interesting and important question to ask, but I think it's really the wrong question to ask.
We shouldn't base our behaviour in the normal case, on addressing the rare, possibly non-existent, hypothetical case. Overattention to rare or nonexistent hypotheticals is a common pattern in human behaviour and leads to very many serious evils. It's morally wrong to subject 2.75 million air travellers per year who are not terrorists to invasive searches just because of the hypothesized existence of three or four who may be terrorists - even if those invasive searches would actually stop some terrorists. (It's clear they don't anyway, but that's another story.) It's morally wrong to violate the property rights of persons who buy DVDs - the usual case being that we buy them to watch them - just because of the fear we might pirate them commercially in the unusual hypothetical case. It's morally wrong to spread hatred and fear against all male human beings because of the rare occurrence of stranger rape. Don't even get me started on the situation of "child pornography" law in Canada. In the lower-stakes realm of software engineering, it's bad practice to spend weeks shaving milliseconds off your garbage collector if doing so costs you microseconds on the normal-course memory accesses that occur millions of times more often than garbage collection, so you lose overall. The usual and routine cases really matter. This is a general statement of the legal aphorism that "Hard cases make bad law." We need to design for what's normal - being aware of other things that can occur, but not allowing them to eclipse correct handling of what's normal.
What would happen if I mysteriously vanished? Well, that's an extremely unlikely hypothetical situation. It's not actually going to happen. What I do in the normal course of events should not be hostage to extrapolations of what might happen in a situation that won't actually occur. Looking at the actual probabilities, by far the most significant reason for me to suddenly vanish from the lives of the people around me would be if I chose to do so; if you'll pardon my being both macabre and blunt, the statistics suggest that for persons of my age and medical history with my lifestyle living in the place I do, if I were to die in the next year, by far the most likely cause would be suicide. (If you don't like thinking about that, it may be lower-impact to think about the idea that I could move to a different city; it is known that I've done that several times already.) The point is that my disappearance for reasons outside my control is not part of the reasonably expected list of situations I should be basing my normal-course decisions on.
What does matter isn't the hypothetical situations that won't happen, but the ordinary situations that do routinely happen. What happens on an ordinary day when I do show up and have not vanished? Don't tell me how much I'd be missed if I were gone, and tell me to plan for that to be as much as possible. It's not going to happen; it is not real. Instead, it makes more sense to look at how people appreciate me in the normal course of events; the normal course of events happens nearly all of the time. How much people appreciate me when I'm present, not how much they would miss me if if if I were not present, should be the basis for decision-making. What happens when I show up on every ordinary day has the most weight in choosing where I want to show up on every ordinary day.