It is widely believed that grain size reduction by dynamic recrystallization can lead to major rheological weakening and associated strain localization by bringing about a switch from grain size insensitive dislocation creep to grain size sensitive diffusion creep. Recently, however, we advanced the hypothesis that, rather than a switch, dynamic recrystallization leads to a balance between grain size reduction and grain growth processes set up in the neighborhood of the boundary between the dislocation creep field and the diffusion creep field. In this paper, we compare the predictions implied by our hypothesis with those of other models for dynamic recrystallization. We also evaluate the full range of models against experimental data on a variety of materials. We conclude that a temperature dependence of the relationship between recrystallized grain size and flow stress cannot be neglected a priori. This should be taken into account when estimating natural flow stresses using experimentally calibrated recrystallized grain size piezometers. We also demonstrate experimental support for the field boundary hypothesis. This support implies that significant weakening by grain size reduction in localized shear zones is possible only if caused by a process other than dynamic recrystallization (such as syntectonic reaction or cataclasis) or if grain growth is inhibited.