Owing to their high thermodynamic efficiency, pulsating combustion cycles have become an attractive option for future gas turbine designs. Yet, their potential gains should not be outweighed by losses due to unsteady pressure wave interactions between engine components. Consequently, the geometric engine design moves into focus. Ideally, one would quickly test several different principal layouts with respect to their qualitative behavior, select the most promising variants and then move on to detailed optimization. Computational fluid dynamics (CFD) appears as the methodology of choice for such preparatory testing. Yet, the inevitable geometric complexity of such engines makes fully resolved CFD an arduous and expensive task necessitating computations on top high-performance hardware, even with modern adaptive mesh refinement in place. In the present work we look at the initial flow field of a shock generated by a pulse detonation combustor (PDC) which leaves the combustion chamber and enters the plenum. We provide first indicators, however, that overall mechanical loads, represented by large-scale means of, e.g., mass, energy, and momentum fluxes can be well estimated on the basis of rather coarsely resolved CFD calculations. Comparing high-resolution simulations of the exit of a strong shock from a combustion tube with experimental schlieren photographs, we first establish validity of fully resolved CFD. Next we compare several integral quantities representative of overall mechanical loads with a sequence of successively coarser grid simulations, thereby corroborating that the “quick and dirty” coarse-grained simulations indeed allow for good order of magnitude estimates.