Inconsistence of super-luminal Cern-Opera neutrino speed with observed SN1987A burst and neutrino mixing for any imaginary neutrino mass

We tried to fit in any way the recent Opera-Cern claims of a neutrino super-luminal speed with observed Supernova SN1987A neutrino burst and all (or most) neutrino flavor oscillation. We considered three main frame-works: (1) A tachyon imaginary neutrino mass, whose timing is nevertheless in conflict with observed IMB-Kamiokande SN1987A burst by thousands of billion times longer. (2) An ad hoc anti-tachyon model whose timing shrinkage may accommodate SN1987A burst but greatly disagree with energy independent Cern-Opera super-luminal speed. (3) A split neutrino flavor speed (among a common real mass relativistic neutrino electron component and a super-luminal neutrino {\mu}) in an ad hoc frozen speed scenario that is leading to the prompt neutrino de-coherence and the rapid flavor mixing (between electron and muon ones) that are in conflict with most oscillation records. Therefore we concluded that an error must be hidden in Opera-Cern time calibration (as indeed recent rumors seem to confirm). We are also reminding the relevance of the guaranteed minimal atmospheric neutrino mass whose detection may be achieved by a milliseconds graviton-neutrino split time delay among gravity burst and neutronization neutrino peak in any future SN explosion in Andromeda recordable in Megaton neutrino detector.