It reminds me of my days as a general surgery resident, when one of the most common conditions leading to surgery was stomach ulcers. (They are prone to bleeding and numerous other unfortunate consequences.) Standard advice regarding ulcers was to avoid spicy foods and alcohol, in the belief that these encouraged the stomach to produce more acid. It was some years later that we learned that the culprit in most cases was a type of bacteria, which as luck would have it is inactivated by wine polyphenols. So ironically we should have been encouraging these patients to have wine with dinner, contrary to all prevailing logic.
The real point here is that studies of this type, called population studies, require a healthy degree of skepticism in order to derive anything useful. For one thing, they rely on self-reporting of drinking habits, which is notoriously unreliable; for another, there are few populations with consistent drinking patterns anymore. The best information actually comes from studies done years ago when such populations did exist, for example in rural France. These studies consistently find health benefits to regular moderate consumption of wine. The final lesson here is that regular moderate wine drinking is linked to other healthy behaviors, so that it can't be reduced to a chemical formula and put into a pill. So when you see the ads for the newest supplement touting "all the benfits of wine without the alcohol" you know it just ain't so.