That first one is really interesting. The second just talks about reduced alcohol beer, and it's a lot harder to tell the quantity of alcohol by smell/taste than the presence of alcohol. But swapping something out with a beverage with no alcohol is a completely different story, so I have to assume it's because it's in a social atmosphere where other alcoholic beverages are present, so smell isn't reliable and taste is secondary (they're focused on the conversation, not the drink).

even sommeliers get fooled easily

Yeah, I've read/seen studies about just that, and I think it's hilarious that the main difference between an expensive and cheap wine is branding. But distinguishing drinks by taste is different from determining whether a drink has alcohol.

As an anecdote, my brother went to a wedding reception while underage and there were two bowls of juice, one spiked and one not. He accidentally got the spiked punch, tasted it, and wondered why it tasted funny, at which point my parents noticed he got juice from the wrong bowl. He didn't know what to look for (he was like 10 at the time), but he knew something was off.

So I think if you take away the "distractions," it'll be very easy to tell whether a drink is alcoholic. But maybe I'll give it a try and do a blind test. I don't think I'll even need to get to the tasting part, the smell alone should be enough to distinguish alcoholic from non-alcoholic beverages.