To make it even more confusing, there are at least dozens of strains/families of each culture, and the strain you get depends on the manufacturer. For example, for L. Helveticus, not all strains help significantly with bitter peptides, it's actually a select 5-10 according to studies. Similarly for diacetylactis, the rate of mg diacetyl produced per liter varies among strains. And if you use a mother culture, then that rate can change over time with mutations, bacteria adaptations, colony dominant strain dynamics, etc.
I really like your wine analogy because the biochemical complexities are similar. With wine you have all sorts of complexities like dissolved O2 for kreusen/initial yeast growth, available N and minerals for yeast cell wall strength to resist osmotic pressures, as well as initial tannin extraction through enzymatic and yeast action, and then during aging the rate of oak compound extracted, and how it interacts with tannin and SO2 levels, evaporation in the barrel, and thousands of other issues.
I think cheese is actually more straightforward