Coulomb Blockade with Dispersive Interfaces

What quantity controls the Coulomb blockade oscillations if the dot--lead conductance is essentially frequency--dependent ? We argue that it is the ac dissipative conductance at the frequency given by the effective charging energy. The latter may be very different from the bare charging energy due to the interface--induced capacitance (or inductance). These observations are supported by a number of examples, considered from the weak and strong coupling (perturbation theory vs. instanton calculus) perspectives.