Numerical modeling and simulation of multiphysics coupled systems is still a potentially fruitful field of research, even when it comes to intrinsically linear or linearized formulations. This is usually the case with computational models of vibroacoustics coupled systems, where it is not uncommon that the combined effect of many localized geometrical missmodeling with significant uncertainty in mechanical characterization of some materials - typical for organic fibers, for instance - propagates to large discrepancies in the natural frequencies and mode shapes obtained from ordinary finite element analysis. The main goal of this work is to compare two basic approaches for the modeling of stringed musical instruments in the frequency domain: finite element modal analysis and simplified lumped-parameters analytic modeling considering only the most influencing degrees-of-freedom (accordingly yielding the first few resonances). Thus, the focus here is to review some important references in this topic, including previous work from the authors which is expected to bring useful alternative ways of addressing some specific problems. Albeit restricted to the slowly varying structural behavior, simplified analytic models proved well capable of closely representing the dynamics of the lower frequency range, while still amenable to the identification of its parameters through simple routines relying on experimental data. On the other hand, properly updated finite element models allowed long-sought connections between physically relevant values and crucial subjective quality descriptors - essential pieces of musicians' and luthier's vocabulary - to be eventually found.