Yeah but you have to actually do a study to determine that relationship, which is exactly backwards from what this "study" did.

Flu rates are a much simpler metric than "most stressed," which is a very squishy phrase that they don't even bother to define. Afaik no one tracks stress levels nationally, and if they do, there's a million different ways that'll give you a million different answers. "How stressed are people" is a very different question from "how many people have the flu." Even if they wanted to do this right, it would be a lot harder because the CDC can't just give you a number for you to correlate with search data.