As any of us has dabbled in C-41 processing knows, capacity is all over the brand ball park. Many say 8 per qt/liter. Then there is the matter of increased time to extend capacity
There really is not much leeway here with respect to the process specs; the so-called capacity is really more of: "how far off is it allowed to get before it bothers you?" So Kodak, as the designer of the process, takes a fairly conservative approach; some of the aftermarket suppliers rely on the frugal nature of their customers, who seem to not be bothered, or even notice, the difference.
"Perfection is the enemy of good enough."
Yes, I think the main thing here is, what does "good enough" mean?
David hasn't said what sort of thing he photographs nor how finicky he is, so it would be hard to judge. A lot of things, especially scenic shots, have a lot of leeway for "interpretation;" how can one look at a landscape and say, "oh the change in color of foliage looks like a color cross?" But if you shoot certain sorts of things, say products, or one of my specialties, studio portraits, this can be real touchy. Depending on how finicky you are. If you were perfectly satisfied with the quality of mini-labs in their heyday, circa 1990s, then you're unlikely to notice color crosses; good quality portrait work is whole 'nuther animal. (The place I worked also owned a large 1-hour lab chain; they set up a number of attached portrait studios as a trial, but could never get what I'd call high quality. Note that I had some involvement in this.)
The rest of this post details some of my experience with the portrait chain, going into how we tested for film/paper compatibility. No need to read further unless you have specific interest. It's mainly to demonstrate how some users can be more finicky.
I've spent a lot of years working with high volume processing for a large portrait chain. On the general topic of color crosses, this was a major issue for us for many years. As Kodak pro color neg films, the only thing we ever used, went through different generations - VPSII, then VPSIII, and the first Portra 160 - we would thoroughly test each one, printing one the appropriate professional paper. Since portraits were our business, this was a major part of our testing. We'd shoot a wide range of exposures (with a large number of studios, you always have someone screwing up the exposure) using models with a variety of complexions and hair color. Then they were all hand color-balanced (optical printing) to match the skin tones to within about 1cc color. (This testing regimen went on until about a dozen years ago, by which time we were fully digital.) One main requirement was that color crosses across the skin tones - shadow to highlights - had to be minimal. A common problem in the earlier years, depending also on the paper used, was that the brightest skin highlights would go "cool," meaning a bluish or cyan tone. In later years, even across paper brands (still using the pro portrait papers, though) these color crosses were minimal - on the order of 1 cc or less. Which we judged that no normal customer would see.
Now, when we did these tests, all of our processors were pretty well nailed down with respect to control strip plots. So I truly do not know how the test result would be with an out-of-spec film process. But I do have a sense of what the control plots will do if when the temperature control goes off, or a replenisher pump fails. The "speed" between the color layers (c, m, and y) will shift, as well as the relative contrast levels. So I am pretty certain that the grey scales would not remain dead neutral. The skin tones are not "predicted" by control strips, as skin doesn't have a neutral balance on the exposure, like the control strips do. So the best way I knew to test the skin tones for color crosses was to shoot and print it. If someone has not done this I don't know how they could possibly anticipate what it would be like. (Today, with scanning the standard method, it's a different sort of game; I can't say too much about it.)
Anyway, given the difficulty we had, with the best materials and tight process controls, I cannot imagine that someone can pick some arbitrary dilution/time/temperature and match the results. My best guess is that David has more, uh, relaxed standards, or shoots subjects that don't really demand a specific sort of color reproduction - one where color "errors" don't really stand out.