In their paper, Anshul Saini and Dejan Stojkovic  claimed that by calculations it is possible to demonstrate that in a gravitational collapse of a disk, an event horizon is never made for a far observer. Some "radiation" is emitted in the moment of collapse and when the collapse gets close to forming a black hole the "radiation" is very similar to Hawking radiation. However, the emitted radiation "turns down" the collapse in such a way that as time goes to infinity we get only "radiation."
I'm concerned about the generality of this result. Does this mean that black holes are some sort of approximation for small time intervals (when compared with the mass), in analogy with any other resonance? For example, the eigenvectors of a hydrogen atom of $1p^2$ is a resonance via interactions with EM vacuum. Is the black hole an analogue of a $1p^2$ state in this sense?
 Radiation from a collapsing object is manifestly unitary, Anshul Saini, Dejan Stojkovic (SUNY, Buffalo), Phys.Rev.Lett. 114 (2015) 111301