It has been answered several different ways. If you don't find them "acceptable," that's your problem. Truth doesn't care what you don't like to hear.
Now, I don't hold the counting from midnight view myself, but you're insisting on something you haven't proven yet. Are you a historian? Do you have formal training? Are you an armchair enthusiast of ancient history who happened to read a few journal articles on ancient timekeeping? Can we see your sources? Are they primary, or at least peer reviewed secondary sources? Or are you simply insisting on the absolute and unconditional veracity of something you read once upon a time somewhere?
And again, have you ever been long separated from a wristwatch?
You want timekeeping to be absolutely precise, but also count hours from dawn. Do you not realize that these two propositions are mutually exclusive? To count time exactly, you need some kind of exact measurement system for the passage of time. That system has to be calibrated to some kind of observational benchmark that always occurs at the same time. The only such benchmark is high noon. Sunrise occurs at wildly different hours over the course of the year. If the hours begin at dawn then the hours can't have fixed length and timekeeping is nothing but estimation.