Think of it this way: if 10% of every other generation does X, but 40% of a particular generation does X, then that generation really should be known for doing X, even though the majority of members of that generation don't do X! Don't let the easy availability of anecdotal counterexamples to a trend fool you; a trend in a population is a trend even if not literally everyone in that population is a part of it.
Of course, there might well be a better way to divide up the population that serves as a better predictor for the trend, and so "explains away" the generational effect. E.g. changes in racial composition (and so culture) due to immigration over the years; or growth of subcultures as self-reinforcing memetic entities, apart from any generational proclivity toward them; or economic effects pushing people to have different emotional needs (e.g. a change in gender-balance of the workforce will see different types of businesses created to offer self-care to tired workers.)
But, by another lens, all these other effects are "part of" the generational effect; they're the composition of the narrative the generation will tell about itself, and so be remembered for.
