Nonlinear device modeling of high-power microwave transistors has a rich history of incorporating the electronic behavior with temperature effects as these devices operate at high-frequencies and high powers and dissipate a significant amount of heat. The sophistication of modeling approaches has increased significantly with increased computing capability and availability over the past several decades. With this we see a new type of model, one that couples the electromagnetic, thermal, and devices physics simultaneously and in a distributed fashion so that many transistor fingers, or even sub-sections of fingers are modeled. While these new approaches provide significantly more detail into the internal operation of the transistor, they are not without increased difficulty for the modeling engineer. This paper provides further motivation for these models by placing several recent developments in context while demonstrating the additional information and value that a distributed multiphysics-based modeling approach can provide.