There is, I am starting to believe (regardless of whether or not this particular finding holds up), an ongoing problem with our approach to nutrition. The anti-pattern goes like this:
1) people who eat/drink X have health benefits
2) it seems to be related to X having Y in it
3) let's sell Y as a pill
4) dang, no health benefits, how did that happen?
Now this could be for several reasons. Maybe eating/drinking X is a marker for a lifestyle that is healthy, or for coming from a background that includes beneficial genes of some sort. Maybe Y is not the only thing in X which matters, it's just the most obviously related thing, and you really need the whole package. Maybe the amount of Y you get from eating/drinking X is good, but in concentrated form at high levels there are drawbacks of some sort.
But, however you look at it, putting Y into a pill (that can often be patented and sold as medicine) is not the same as eating/drinking X, and despite the financial incentives for claiming it is, we need to recognize that.