Interesting.

Re: "other [variables] are at least as important" being an incorrect assumption...

I can see this being possible for some foods (your salmon steak being a good example, given you include thickness, but you'd need different models for many different types of fish), but really unlikely for others. How onion behaves, for example, is pretty regional and highly dependent on type. For many vegetables water content is highly variable and very important to how it cooks, but I feel like i've got better proxys for what is going on than cooking surface temperature would be....