Carlo,
I agree with you that if one writes a Schroedinger equation one gets in turn unitarity without many calculations (in fact I stressed that my calculations in my Essay are on the same level of university studies on quantum mechanics). On the other hand, the vice versa, i.e. that unitarity permits to ALWAYS write a Schroedinger equation is not so trivial as you claim. I recall you that Maldacena, Susskind and the same Hawking claimed that unitarity is restored in black hole evaporation through the ADS/QFT duality, that Mathur claimed that unitarity is restored through the "fuzzball" approach and that Zhang, Cai, Zhan and You claimed that unitarity is restored through the correlations among Hawking quanta. In any case, neither Maldacena, Susskind and Hawking, nor Mathur, nor Zhang, Cai, Zhan and You wrote down explicitly a Schroedinger equation which permitted them to find a pure final state. In fact, unitarity is also permitted WITHOUT obtaining a Schroedinger equation as in some cases one obtains a final mixed state but information is preserved trough correlations among the subsystems. I did NOT construct the Schroedinger equation by assuming unitarity. I constructed the Schroedinger equation by using my result in Int. Journ. Mod. Phys. D 21, 1242023 (2012), where I have shown that black holes quasi-normal modes can be interpreted in terms of quantum levels. I do not think that such a result automatically implies unitarity, because in that case, I should have won a Prize in the 2012 Gravity Research Foundation Competition rather than a "simple" Honorable Mention as in that case the black hole information paradox was implicitly automatically solved in my Essay Int. Journ. Mod. Phys. D 21, 1242023 (2012)!! On the other hand, you claims that "In the classical theory, there is no time translation invariance at infinity. Information can fall in, instead than out." But here I am not using neither the classical theory nor the full quantum theory. I am using a semi-classical approximation. Maybe time translation invariance at infinity is restored in my result Int. Journ. Mod. Phys. D 21, 1242023 (2012), I have not checked. You can check it if you like.
Cheers,
Ch.