What's more radically different between people who are healthy and people who are not is how much they exercise rather than what they eat. There is much more worldwide cultural variation in what people eat, and much more individual variation within each culture, among healthy people than there is variation in the exercise level of healthy people. But seeking to change diet to change health sounds easier than getting more exercise, so that's what gets in the news (and on HN, over and over and over).

Doing serious scientific research

http://norvig.com/experiment-design.html

with controlled experiments and careful observation and follow-up on human nutrition is extremely hard. I recall a local television news documentary about a dietary study from the 1970s in which experimental subjects were confined 24/7 in a lab for many weeks (they must have been compensated handsomely to agree to participate in this experiment) and fed EXACT portions of food that were weighed with extreme precision. Oddly, I don't recall particularly what hypothesis that study was set up to test, but I do recall that it was described as the first experiment that relied on feeding subjects exactly measured portions of known foods (and excluding subjects from access to foods they chose for themselves) while involving lengthy (weeks of time) follow-up. It's still exceedingly rare for studies of human nutrition to have careful experimental designs. (And even at that, I suppose most of the subjects of that experiment are still alive, but I don't know if anyone is doing follow-up observations of those subjects in any manner.) Instead, we get inferences from Egyptian mummies (as in the submitted article) or from other uncontrolled case studies that never even had careful observation to begin with. That really isn't science;

http://norvig.com/experiment-design.html

that is trading anecdotes, and third- or fourth-hand anecdotes at that.