Optimization of the energy resolution of Josephson tunnel junction X-ray detectors requires the minimization of event-to-event variation in the number of nonequilibrium quasiparticles that tunnel and are thereby detected for monochromatic input. This requires a detailed understanding of the determinants of the time scales for the degradation of the energy and expansion of the disturbed volume, of possible quasiparticle (QP) self-trapping and phonon bottlenecking, and of the impact of materials parameters such as grain size. Moreover, the signals of distributed detectors cannot be interpreted without a detailed model of the evolution of the spatial variation in QP density. These matters, and the extent to which the details of the QP tunneling pulse probes the energy degradation and expansion, are discussed. The expansion is modeled in two dimensions with Monte Carlo simulations of ballistic quasiparticle propagation between randomizing collisions and compared with published data for a distributed junction detector. Alternative explanations of the high quasiparticle loss rate and curvature seen in the published plots of the relative fraction of the QP detected by the two sensing junctions are offered.