After reviewing some of the literature, I stand corrected on two points- oral melatonin does show some effect at reducing sleep latency above placebos, but unlike most prescription sleep aides, it does not affect the depth of sleep (ie, it may help you fall asleep faster but it doesn't keep you from waking up). But because these studies were conducted in controlled, quiet environments, it is difficult to generalize the results to sleeping on a plane.
Second, there is some good evidence (I was not previously aware of this trial) that melatonin, when taken over a period of days, can help re-adjust the circadian rhythm disruptions caused by jet-lag faster than a placebo controlled group. The crucial difference between this trial and previous trials is that neither the treatment group nor the placebo group were exposed to any light above that which they would normally experience in an average given day. The differences between the two groups were not large, but did meet the requirements for statistical significance. Modafinil (brand name Provigil), however, has been much better studied for this and has shown clear superiority over both melatonin and a placebo.