Food: Yes, but that's a different case, because I had lumped food into one mass (which either had poison or didn't) and you had divided onto two plates (one poisoned but not both). I was addressing whether you do something to get rid of the poison; you're addressing whether you eat the nonpoisonous plate of food. The labeling of anything already familiar to us is usually subconscious but not less than a label for being subconscious; it is also usually rapid and not less than a label for being rapid; even combining both subconsciousness and rapidity does not make the signifier less than a signifier.
Color: I was referring to the physical condition, not a cultural one. Human eyes have nerve receptors in the retinas that sense within a color range; most of us have those for three color ranges, with the ability to perceive one hundred shades in each range (combining yields 100 x 100 x 100 = 1,000,000 colors). Estimates are that two percent of women and zero percent of men have genes for tetrachromacity but that of the two percent of women very few biologically express that genetic characteristic, due to epigenetics. It's also likely that culturally we are taught as children not to see additional colors; a tetrachromat's parents probably are not tetrachromats and they'll teach their children how to understand what they see, especially since human children, unlike any members of other species, see pictures in books and on TV and those media don't support tetrachromacity, so the most of the children who do have tetrachromaticity will ignore it, just as we learn in the first few months of life to ignore vocal echoes that aren't useful. While many retailers sell products for the "long tail" (few sales per item across many items), such as for books (Barnes & Noble as of a few years ago reportedly carried many titles that sold at a rate of one copy every 3-4 months per store), I have not seen that effect (in the big city where I've lived for years) for product color choices that would appeal only to tetrachromats, so they're not likely to be anywhere near two percent of the population. One tetrachromat lives near a river and can see, in the river, colors most people don't see, so tetrachromats evidently find some of the additional colors not only in paintings and color swatches (arguably artificial) but also in nature. There is a little bit of research on the subject available online.
Some cultures name only two colors (white and black, or, I think about as likely, light and dark) and there's a sequence for add-ons (if there are only three colors named, most three-color cultures have the same third color, if there are only four colors named, most four-color cultures have the same fourth color, and so on, but not all the way to very large name sets, probably because agreement evaporates near the high end). It's likely that they don't find it useful to distinguish more colors, perhaps because, for what's important to them, other characteristics suffice to distinguish the items. If we need to distinguish between elephants and giraffes, we don't need "gray" and "orange" because there are enough other differences. And even for a culture without a name for 'green', that's a name composed of sounds or writable characters; if they match a green specimen to a palette of greens, either the sample doubles as a signifier and the palette is a palette of signifieds or the palette is a palette of signifiers and the sample doubles as a signified. That's like a child pointing to someone; the someone is a signified and the pointing motion is a signifier.
On Sapir-Whorf: I agree that language influences perception in individuals and cultures, but I'm not relying on that here. Rather, I'm arguing that perception, in this case that a choice is available and what that choice is, influences language and, earlier in life, biocommunication systems that can accept symbols.
The assigning of communication symbols is probably driven by utility. We can tell pet cats apart but most of us, facing a forest, would have a hard time telling oaks from other thick trees, never mind oaks from each other. Some primates, however, know each oak individually. We can look at a valley of grass and enjoy it but we'd hardly give every blade its own name. Few would try, even for fun. But identifying mushrooms would be vital, because some are delicious and some will kill you. And humans are not the only one who had better identify mushrooms. Mushrooms can't run away, so their survival may depend on prolific reproduction or discouraging eating them, and discouraging works much better if prospective diners remember which is which. Prospective eaters include many animals, including birds. Having a bird brain may be a very good thing, noted one ethologist.