Peter makes no claim as to what people choose to do or why they choose to do it. Regardless, people’s beliefs are irrelevant to the question of whether alcohol consumption is beneficial.

Peter only cites “recent evidence,” never appealing to any popular belief. Peter's argument also doesn't make any comparisons.

Peter’s conclusion is simply that alcohol consumption is beneficial, so it doesn’t matter if there are other methods that are too. The issue is assuming that alcohol consumption in general has net-positive effects just because moderate drinking has certain limited benefits.

This accurately describes how Peter’s evidence is too limited to support his conclusion. The evidence only suggests that “moderate" alcohol consumption has certain benefits, yet his conclusion is about alcohol consumption in general.

Notice that (D) doesn't address the "net-positive" issue—but that's okay! This still accurately describes a flaw, even though there's another flaw the correct answer could have targeted.

For alcohol to have health benefits by killing harmful bacteria, it's not necessary that alcohol kill all harmful bacteria. Even killing some diseases is a genuine benefit. The flaws are (1) assuming that this benefit outweighs any potential negatives, and (2) generalizing evidence about moderate drinking to all drinking.