The argument probably is that in various populations certain vitamin deficiencies are so prevalent (e.g. vitamin D in African Americans) that you can pretty muc

The argument probably is that in various populations certain vitamin deficiencies are so prevalent (e.g. vitamin D in African Americans) that you can pretty much assume that a random person is most likely to be deficient in one or more categories and would benefit from a supplement to fix that. Of course, that argument is valid only if the supplement would actually fix the prevalent deficiency, as you refer in your example about vitamin D.