[svat]

Self-reported statements (about how someone feels) are good evidence for how they feel. In your example, if the claim is that your friend believes he is not drunk, then his experience still carries more weight (is a more reliable indicator of his beliefs) than most studies.

This is different from the question of whether he is actually sober, of course. But the post you're replying to is specifically about the claim that X person feels better, not about any physical effects on their body. I'd rather trust a person about how they say they feel (especially at a given moment), than studies trying to determine how most people probably feel or ought to feel in a similar situation.