That was a pretty ambitious statement to make in light of the discovery, but such comments are often made these days. I tend to see that as bordering on the irresponsible. Gene therapy is still in its infancy, and simply finding a gene that seems to be related to something doesn't mean that it doesn't do many other things. If tampering with this gene wouldn't have untoward negative side effects then it would be beneficial, but given that it has just been discovered, the research is a long way from determining all of its functions.
I do schizophrenia research. The majority of people in the field finally have agreed that it's polygenic (several different genes contribute), but that some sort of environmental input, most probably a stress of an undetermined nature, may act as a catalyst. Those initial catalytic stressors also seem to be different for different patients. Several physiological and anatomic markers have been identified, but those represent effect, not cause. Sadly any possibility of gene therapy for this disease is far, far in the future, if it ever will be possible.
As to your title question, I think that it's simply a matter of social definition based on social considerations, not any medical criteria. Alcohol abuse can have such obvious, devastating consequences for the people around the person, whereas the serious effects of smoking on others are more distant in time (except for the noxious fumes themselves, cough!). It also may have something to do with the comparative (eventual) willingness of an alcoholic to accept treatment, if the condition is described as a disease as opposed to a behavior voluntarily undertaken. Both are valid.
