Chewie, while a generalization is not applicable to any single individual, they can be useful when discussing the behaviors of groups as a whole. So, pereception/stereotype, generalizations, and statistical proof or disproof of each can and should be taken into account when discussing social phenonmena.
"As good sociologists, we make generalizations about groups, not individuals." -- Berger.
So, while they may very well sound like stereotypes, social facts such as "Japan has the highest suicide rate" or "African Americans drink more Kool-Aid per capita than other ethnic groups" are quite relevant in a social study context. I try to limit my uses to those that are fair social patterns, and I would of course suggest everyone else do the same.
But some stereotypes also happen to be true.
Of the group of course -- not a particular individual.