Good question. The paper says 0.1 mmol per liter, i.e. fruit is sprayed with a solution containing about 23 milligrams per liter of melatonin.
It's possible that the amount retained would be less than a few milliliters of solution equivalent — less than 100 micrograms under practical consumption patterns — so having little activity. But if the melatonin is absorbed by the fruit, the effect could be significant and particularly problematic since people often eat fruit in the morning.
In any case, I think that a more likely goal of this research (vs actually spraying commercial fruit with melatonin) might be to identify compounds able to produce the same beneficial effect on fruit without causing any adverse effects in humans.