Thin-film composite reverse osmosis (RO) membranes have dominated desalination technology for decades, with the solution diffusion model (SDM) serving as the gold standard for interpreting mass transfer phenomena since its inception. However, debates still persist regarding the effect of charge on separation performance, owing to the SDM's inability to explore membrane charge. In this study, polyethyleneimine (PEI) was utilized to modify the surface of a commercial seawater desalination membrane (SW30, Dupont Company), imparting regulatable charge properties based on the solution pH. The separation performance of the modified RO membrane was experimentally validated under varying NaCl feed salinities. Specifically, the Donnan-steric pore model with dielectric exclusion (DSPM-DE), known as a charge-responsive model, was employed and juxtaposed with SDM to authenticate the experimental findings. The results indicated that the dielectric effect primarily governed ion distribution (Na+ and Cl−) within the membrane, followed by the steric effect, and ultimately, the Donnan effect. Consequently, DSPM-DE aligned well with the observed salt rejection only when the membrane was strongly charged (that is, under acidic or alkaline conditions), but performed poorly in the neutral state. This observation, rarely revealed in previous research, stems from conventional desalination membranes being negatively charged in most natural water environments. Conversely, the SDM aligned well with the experimental data because it solely considered the solutes' solubility parameters and diffusion coefficient, independently of the surface charge of the desalination membranes. Given the extensive research on surface modification, our study demonstrates the limited influence of charge effects on the selective enhancement of RO membranes.