Is it not far enough to cross the Atlantic? The top end of that estimate would cover that, and there are anedoctal reports of storms bringing coconuts from Africa to South America [citation needed].

Possibly. In my head, I picture it as a kind of exponential decay: more days at sea increase odds of seed failure; slowly at first, then more rapidly. But given enough trials, a rare event can occur at the extremes of probability. Storm events probably skew that distribution.