The legacy of solar neutrinos suggests that large neutrino detectors should be sited underground. However, to instead go underwater bypasses the need to move mountains, allowing much larger wateř Cerenkov detectors. We show that reaching a detector mass scale of ∼ 5 Megatons, the size of the proposed Deep-TITAND, would permit observations of neutrino "mini-bursts" from supernovae in nearby galaxies on a roughly yearly basis, and we develop the immediate qualitative and quantitative consequences. Importantly, these mini-bursts would be detected over backgrounds without the need for optical evidence of the supernova, guaranteeing the beginning of time-domain MeV neutrino astronomy. The ability to identify, to the second, every core collapse in the local Universe would allow a continuous "death watch" of all stars within ∼ 5 Mpc, making practical many previouslyimpossible tasks in probing rare outcomes and refining coordination of multi-wavelength/multiparticle observations and analysis. These include the abilities to promptly detect otherwise-invisible prompt black hole formation, provide advance warning for supernova shock-breakout searches, define tight time windows for gravitational-wave searches, and identify "supernova impostors" by the non-detection of neutrinos. Observations of many supernovae, even with low numbers of detected neutrinos, will help answer questions about supernovae that cannot be resolved with a single highstatistics event in the Milky Way.