Former chef here, the variance in balances of different chemical constituents can be so massive between two pieces of fruit picked from the same tree that precise recipes would be useless without requiring each reader to have access to a mass spectrometer (and any clue as to what its readout means). This is largely where the "art" in culinary arts comes in. While you could teach someone to read a mass spectrometer to determine the exact chemical result that a given mixture of different ingredients applied to a specified heat curve, it's more likely that they'll just have to learn how to approximate it via taste, smell, and sight.
Good baking books will actually get a lot closer to having specifics in them. They can do this however because a lot of the ingredients that are used in baking are rated based off of various levels of one or more of the chemicals that will significantly affect the result of your baked product (ash level, acidity, sugar content).
This raises an interesting question, relevant for more than cooking: In a process with variable "inputs" , is there a method(probably statistical) to create a recipe that will work,even with this variability, while minimizing the expertise needed ?
"Time Tested" recipes have already done this work - they've performed the process over a long period of time with a wide variety of small input changes/variable ingredients.
For example, the Toll House Cookie recipe has been executed how many millions of times with predictably good results.
I’ll answer for him. It’s a terrible, stressful, dangerous job, with insane hours & criminal working conditions, that pays peanuts. Might just be my opinion.
Good baking books will actually get a lot closer to having specifics in them. They can do this however because a lot of the ingredients that are used in baking are rated based off of various levels of one or more of the chemicals that will significantly affect the result of your baked product (ash level, acidity, sugar content).