Is that necessarily true? Film has a limited latitude. I can envision a scenario where you take a reading off a grey card, but the detail in the snow is so far (more than 4 stops) above that that you can't actually capture any detail of the snow.
It is really true that, assuming perfect calibration with your meter, the grey card would end up middle grey, and that the luminance range of the objects in the composition would fall around middle grey. As you reasoned, if the edges of this range land outside of your desired range, they become too dense or thin to make a rendition that perfectly matches what you want (when the grey card is printed to middle grey on the base paper; grade 2 or filter 2-1/2, for instance).
So, the grey card/incident meter basically does one main thing: it tells you how to render a grey card as middle grey, and therefore get everything roughly into place tonewise. It cannot always match the luminance range of the scene to the desired black to white range on the print. This is where an eye for contrast in the composition, or a spot meter, can help you. The way you expose and develop will let you control the contrast of the scene. If you know that a scene will end up too contrasty if shot on your film, you overexpose the film to give it more density across the board. Then you underdevelop it in order to maintain this density in the shadows, while bringing the highlights back to where they would have been if the shot had not been overexposed. You can do the opposite to increase contrast by making the shadows thinner, while pushing the midtones and highlights back to where they would have been if they had not been underexposed. You can also overexpose-overdevelop, or overexpose-normal develop to increase the contrast between the darker tones, while either pushing highlights farther, or keeping them more in check, respectively.
So, in a scene that has an easily printable luminance range (when your scene has a 10 EV split between what you want maximum black to what you want paper white), you are fine going what the meter/gray card say, developing normally, and printing on your normal grade or filter. You would be fairly easily printable if you were a step low in contrast or a step high in contrast. You would be printable, though challenging, if you had to go up or down two filter grades. You would be printable as a last resort if you had to go up or down three paper grades.
I think of it like this. There are three basic scenarios when metering ambient light in a composition:
1. What you want to be maximum black and paper white in the composition are about 10 EVs apart when measured at the scene. The darkest and lightest areas in which you want to see texture are about 6 EVs apart. The darkest areas in which you want to detail are 4 EVs apart. This is a "normal" luminance range for your paper to match.
2. What you want to be maximum black and paper white in the composition are greater than 10 EVs apart when measured at the scene. The darkest and lightest areas in which you want to see texture are greater than 6 EVs apart. The darkest areas in which you want to see detail are greater than 4 EVs apart. This is a higher than normal luminance range for your paper to match.
3. What you want to be maximum black and paper white in the composition are fewer than 10 EVs apart when measured at the scene. The darkest and lightest areas in which you want to see texture are fewer than 6 EVs apart. The darkest areas in which you want to see detail are fewer than 4 EVs apart. This is a lower than normal luminance range for your paper to match.
...and then think of it like this:
Knowing through judgment, or metering, how close to situation one you are is what you need to be thinking about when using that grey card (or incident meter). Once you know this, then:
1. In situation one above, expose as the meter recommends from the grey card reading and develop as normal.
2. In situation two above, overexpose the grey card by as many stops as the luminance range is above the desired print range, then develop for less time than normal (until your highlights land where they would have been if you had exposed and developed for situation one).
3. In situation three above, underexpose the grey card by as many stops as the luminance range is below the desired print range, then develop for more time than normal (until your highlights land where they would have been if you had exposed and developed for situation one).
If you do that, your negatives will land much closer to the contrast filter (or grade) paper you would use in situation one.
All of this works best only when you have determined a working EI and normal development. Exact pluses and minuses would help a lot, but can also be guessed at fairly well as long as you have a normal time.
In other words, it is just a slightly different approach to classic zone system theory (or vice versa). The grey card/incident way, you always meter how to make middle grey middle grey, and use luminance range as the factor that decides exposure and development. The classic zone system way, you place a low tone to make a metered shadow tone look how you want it to look, then you place a metered highlight where you want it via development. In the classic zone system method, the thought process behind it is that one is swinging the characteristic curve up or down from its toe, like a whip. In the grey card/luminance range method, your thought process is that you are seesawing the curve around middle grey. Pick whichever one of these two basic though processes makes more sense for you while working in "the field" (or neither, or both, or variations that work for you).
(Personally, I prefer the classic zone system thought process when I can develop exposures individually, such as with sheet film, or when I have interchangeable magazines, each of which is marked with different development instructions. I prefer the grey card/incident thought process when shooting shots that cannot be developed individually. With hand held shots, I will generally estimate an average luminance range for the lighting in which the roll was shot. With tripod shots, I usually measure the luminance range with a spot meter.)
, but I was trying to show that if you simply metered a snow scene using your camera meter, it would record as 18% gray, not the white you expected. Whereas metering off the gray card would get you a more realistic image and the snow would be "white" as compared to muddy gray.
