The idea that there are four basic tastes—sour, salty, bitter, and sweet—was widely accepted until 2002 when the taste receptors for glutamate were discovered which gives food the flavor called “umami”. (Bacon, parmesan cheese, soy sauce, and tomatos have lots of it.)
But of course this 5-taste model describes only the tastes detected by the tongue. Most of the flavors we identify in food come from aromas. Some research suggests that the average human can distinguish millions if not trillions of distinct odors, some of which emanate from food. So the range of flavors we can detect is quite large.
It is then strange that taste sensations but not flavor sensations are used extensively as metpahors. We routinely use taste sensations to describe emotions, personalities, facial expressions, etc. A person is a sourpuss, a smile is sweet, resentment is bitter, language is salty. But we don’t describe persons as fragrant, minty or herbal. It is curious why taste rather than aroma is the source of metaphorical association.
But at any rate, now that umami is officially a taste I suppose it will eventually acquire metaphorical associations.
So what is it like to feel “umami”?
Exploring the Philosophy of Food and Wine 
I disagree with your claim that we don’t use a wider range of flavor predicates metaphorically for persons.
We certainly don’t use elements of the wider range (i) as frequently, and (ii) with such a constrained range of metaphorical interpretation, as goes for the smaller subset of ‘sour’, ‘salty’, ‘bitter’, and ‘sweet’. But both are predictable: (i) because you would expect frequencies to go down when you go to a larger set of options, and (ii) because metaphorical interpretations crystallize (cf. dead metaphors like ‘mouth of a bottle/river’, ‘kick the bucket’) with conventionalization, which is again related to frequency.
So, I claim, you certainly can apply flavor predicates metaphorically to persons or their personalities (e.g., ‘She has a lemony disposition’, ‘his thinking was minty’, ‘her herbal demeanor’); only the resulting interpretations will be live/interesting rather than dead/boring.
Hi Jonathan,
I agree. I probably should have qualified the claim to be about conventional (or dead) metaphors. There is nothing about aroma metaphors that would make them inappropriate or nonsensical. But it does seem to me that few aroma metaphors have become conventions–or maybe my memory is not functioning well this morning.
What an interesting observation. Umami to me sounds down to earth and stimulating. ‘It was an umami conversation.’
Thanks Foxress. The difficulty with umami is that it is hard to describe what it is like and hard to clearly detect. I’m not sure I could pick it out in a blind tasting at least not without some practice.