[fasterik]

Most claims about whether something is "healthy" or "unhealthy" rely on single studies with small sample/effect sizes. We know that many studies fail to replicate, and in the absence of preregistration there are researcher degrees of freedom for scientists to find whatever they want to find in the data.

I don't know about the current status of alcohol research, but unless there is a large body of replicated work supporting a claim, we should be skeptical of it.