Paths of unitary access to exceptional points

During the early history of unitary quantum theory the Kato's exceptional points (EPs, a.k.a. non-Hermitian degeneracies) of Hamiltonians $H(\lambda)$ did not play any significant role, mainly due to the Stone theorem which firmly connected the unitarity with the Hermiticity. During the recent wave of optimism people started believing that the corridors of a unitary access to the EPs could be opened leading, say, to a new picture of quantum phase transitions via an {\it ad hoc} weakening of the Hermiticty (replaced by the quasi-Hermiticity). Subsequently, the pessimism prevailed (the paths of access appeared to be fragile). In a way restricted to the quantum physics of closed systems a return to optimism is advocated here: the apparent fragility of the corridors is claimed to follow from a misinterpretation of the theory in its quasi-Hermitian formulation. Several perturbed versions of the realistic many-body Bose-Hubbard model are chosen for illustration purposes.