I'm pretty sure what the article says isn't incompatible with Cialdini's Influence at all. IIRC he also says that the more publicly an individual takes a position, the more they feel the need to remain congruent with that position in future. When people decide they want to be seen as a certain kind of person, they want to remain consistent with that. That's why if you want to reach a goal, you should tell as many of your friends and family about it as possible so you risk "losing face" if you fail.
I see this article as basically working the same way. Brands advertise "hey, if you're a chilled out guy drink this beer". Someone decides they want to be seen as a chilled out guy, so they go and buy that beer to reinforce the image they want to convey.
> First off, advertisements definitely do work via simple association... A warmth appeal like a Coke ad with smiling faces will reliably create an association between Coke and positive emotions just because that's how humans are wired.
That sounds like if I was bombarded of pictures of people smiling drinking coke I might forget that it tastes like crap and want to drink it.
So perhaps the social experience someone has when they turn up as the "chilled beer guy" is more valuable to them than the other fuzzy associations the advertiser tries to convey - because the advertiser is making associations that don't have much impact compared to the impact of a real life, human interaction. That's what the guy is talking about when he basically says that the flaw of the 'association' argument is that there's insufficient feedback, only visual cues rather than, say, receiving a massage.
So to summarise I think you're overly confident regarding the power of association by ignoring the social proof aspects and consistency of self image that come with knowing that other people have the same brand associations. I think it's entirely plausible that someone would use those associations as a short cut to convey the self image they want to.