That is all extremely interesting Raedwulf. I know very little about epidemiology, but one wonders if other healthy/unhealthy lifestyles have been factored in/controlled for.
How can they accurately assess life-expectancy without taking into account:

-lifestyles during the childhood and youth of a subject
-diet, in all its complexity, over the entire life
-genetic factors
-environment (pollution for example)
-stressors (life-affecting events, such as redundancy, divorce, bereavement etc)

Surely those and other influences would have some effect on heart health and longevity? So how do they separate those items from the effects of alcohol alone?

Also, I notice that 'temperance organisations' were involved in conducting the study, which is rather like asking the Vatican to study the effects of atheism on populations!