1) Nothing is homogeneous. Even a single static object is not 100% the same now compared to how it was 10 days ago. But in order to make any observation, one has to generalize to the point that has statistical significance.
2) Despite saying Europeans and Americans such articles don't mean ALL Europeans and ALL Americans. For Europeans they mostly mean Western Europe: Germany, France, Switzerland, Netherlands, Belgium, Austria, etc. For Americans they mostly mean the stereotypical middle class suburban American, not a Manhattanite yuppie, nor some Idaho farmer, or "white trash".
3) The idea is not to see if the article's observations hold 100%, but whether they hold more often than not.