<p>Gas transport through nanochannels is critical to applications ranging from gas separation to shale gas extraction. Accurately predicting flow rates under non-adsorbing or weak-physisorption conditions requires the generalized Knudsen theory, which accounts for the full spectrum of gas-wall scattering mechanisms. However, its broad validation across diverse gases and surfaces remains incomplete, largely because the key parameter quantifying scattering, the diffuse reflection fraction (<i>f</i>) or tangential momentum accommodation coefficient (TMAC), is notoriously difficult to determine. Here, we employ molecular dynamics (MD) simulations to develop a robust approach for calculating <i>f</i>. Using this method, we systematically validate the generalized Knudsen theory for a wide range of gases, including all noble gases and common polyatomic species (H₂, O₂, H₂O, CO₂ and CH₄). Our results show that <i>f</i> or TMAC is predominantly a surface property: it clusters around 0.1 for graphene and 0.8 for SiO₂, with only weak dependence on the gas species. By incorporating these <i>f</i> values, the theory achieves remarkable accuracy in predicting flow rates, as confirmed by direct MD simulations across all tested gas-channel combinations. This work establishes a general and predictive framework for gas transport in nanochannels, paving the way for the rational design of nanofluidic devices and advanced separation membranes.</p>