Example: photoelectric effect. The classical description cannot explain (among the other things) the very low time delay between source switching on and photoelectrons detection, because of energy conservation principle. But we know that energy is conserved on average in time; are we totally sure it's also conserved for every instant of time? Or this is a non-written postulate?Energy conservation is ALWAYS respected. The observed momentum values will respect the demands from this law. Besides, why do you think this is not the case ?
You mean: "because the sent photon can go somewhere else than in B?" If this were the meaning, I agree, but it was not this that I intended. What I intended is: do we say that a photon was sent just because it's detected at some distance from the source after have switched on it, or because we can be sure to have sent a photon, before having detected it at some distance from the source? What I "see" in my mind is just an electromagnetic field generated from the source, sent in the form of waves, and I "see" photons as quanta of the energy exchange, only in the interaction between the EM field and the detector. No flying particles (at least no point-like ones).Well, if we observe it at the detector B. QM does not state that there is absolute certainty this will happen each time you send out a photon. Look at the probability distribution at the detector. That's the entire point of the double slit experiment.