Tasting versus drinking, and confidence in ratings
As a 'professional' (yeah!), I make many of my tasting based on a relatively quick sniff, slurp and spit. I'm comfortable with this, because I do it a lot, and I have a reasonable amount of experience.
But as with any data point, it's important to know the confidence you can place in the rating. How much variability might there be in judgements made on different occasions? Could a 91 on one day be a 90 or 92 on another? Or is the variance larger?
The advantage of drinking, rather than tasting, is that there is less variance in the rating. After I've consumed a reasonable portion of a bottle, I'm happy that I've 'got' the wine in a way that I'm not quite as sure about from a quick taste. It's not that I'm unhappy with my judgments based on tasting; rather, that I'm more certain when I've drunk a wine.
It's good to be honest about these issues, and not try to promulgate some notion of taster infallability.