Sure, these things are subtle — my point was that the numbers who would have perished isn't very large in this case, so that under a broad class of assumptions, one shouldn't take the observed absence of nuclear conflict to be a result of survivorship bias.