Errrm. The fallacy that Confident sobriety = non-addictive drinking behavior still comes back to the definition that some sort of sobriety = some sort of drinking. This is simply an oxymoron, a self contradiction at the simplest level. If anyone drinks, they are not sober. Maybe GT's distinction is more useful in this case. If anyone drinks, they are not abstinent.
If the whole concept of sobriety is confusing and ambiguous, then use 'abstinent'. Then the statement, 'I got too confident in my abstinence so I got drunk' is then exposed for what it really is.