Agreed that it's a fully general counterargument. I endorse the underlying point, though, of "evaluate second order effects of success and failure as well as first order effects," and whether or not that point carries the day will depend on the numbers involved.