We quantify acoustic resistance of the gap between a can-annular combustor and high-pressure turbine, using an experimentally validated time-marching computational approach. Adjacent cans oscillating in antiphase drive an unsteady flow through the combustor--turbine gap, creating a vortex sheet, dissipating acoustic energy, and affecting thermoacoustic stability. We use unsteady Reynolds-averaged Navier--Stokes simulations of two-can sectors to excite the antiphase mode and study two cases: a laboratory-scale experiment, and a realistic industrial gas turbine. Predictions of resistance for the validation case agree with measurements to a root-mean-square error of 8\%, allowing for asymmetry of the real apparatus. Further results show linear dependence of quasi-steady resistance on through-gap bias and streamwise Mach numbers, with slope varying from 1.04 to 1.65 across gap lengths due to two-dimensional mean flow effects. In the industrial case, parametric studies of gap length and trailing edge thickness show resistance scales with a trailing-edge Strouhal number, and that vortex shedding is the dominant mechanism driving frequency trends. Approaching the characteristic Strouhal number of 0.2, vortex shedding acts as a source of acoustic energy and resistance falls to zero. By inhibiting vortex shedding, short gaps, vane clocking and coolant flow are shown to increase resistance at high frequencies by up to a factor of 5. Our findings demonstrate quantitative prediction of gap resistance, where existing analytical models fail, and provide design guidance on sensitivities to geometry and the underlying fluid dynamics in can-annular combustors.