I always spotmeter for the specific film and lighting situation actually at hand, along with a general ideal of expectation in the print. The point is a versatile enough negative relative to those parameters. In a high contrast scene, it is important to bag the correct placement of shadow gradation to the greatest advantage. This will differ between films having dissimilar toe structure.

Yesterday I was working with trees with burn scars against bright backlit sky, yet intermittently diffused by thin clouds. So I chose TMX 100 for its long straight line and deep shadow separation, and placed those shadows on ZII at box speed of 100. Sky texture would be effectively brought under rein using the appropriate contrast filter in conjunction with pyro development. But I certainly didn't want to compromise midtone microtonality by resorting to minus development, one of the weak points in the classic Zone System approach in my opinion. Nowadays there are other options, including better VC papers and also contrast masking technique if necessary.

If I just happened have ACROS film instead, I would have rated the film at 50 to get the shadow differentiation further up the toe, and possibly have placed those shadows up on Zone III instead of II like with TMX. Actually, going clear down to Z I placement works with TMax films if necessary; but it simply wasn't that contrasty yesterday, so I could be a bit more conservative.

The point is, I don't accept any one-shoe-size-fits-all model. The haggard old advice to place all shadows way up on ZIII seems designed for
shoot from the hip exposures without good metering skills, or for films like Pan F with ridiculously long toes. Just wasted real estate on the film curve which has to be compensated for at the upper end by minus exposure, thus smashing all the microtonality in between like a stomped peanut butter and jelly sandwich, turning it into blaah mush instead of a sparkly print.