Porous electrodes are core components that determine the performance of redox flow batteries. Thus, optimizing their microstructure is a powerful approach to reduce system costs. Here we present a pore network modeling framework that is microstructure and chemistry agnostic, iteratively solves transport equations in both half-cells, and utilizes a network-in-series approach to simulate the local transport phenomena within porous electrodes at a low computational cost. In this study, we critically assess the versatility and robustness of pore network models to enable the modeling of different electrode geometries and redox chemistries. To do so, the proposed model was validated with two commonly used carbon fiber-based electrodes (a paper and a cloth), by extracting topologically equivalent networks from X-ray tomograms, and evaluated for two model redox chemistries (an aqueous iron-based and a non-aqueous TEMPO-based electrolyte). We find that the modeling framework successfully captures the experimental performance of the non-aqueous electrolyte but is less accurate for the aqueous electrolyte which was attributed to incomplete wetting of the electrode surface in the conducted experiments. Furthermore, the validation reveals that care must be taken when extracting networks from the tomogram of the woven cloth electrode, which features a multiscale microstructure with threaded fiber bundles. Employing this pore network model, we elucidate structure-performance relationships by leveraging the performance profiles and the simulated local distributions of physical properties and finally, we deploy simulations to identify efficient operation envelopes.