Evaluating Mass Outflow Rate Estimators in FIRE-2 Simulations: Towards a Self-Consistent Framework for Spectral Line Based Predictions

$\require{mediawiki-texvc}$Galactic outflows shape galaxy evolution, but their mass, energy, and momentum transfer remain uncertain. High-resolution spectroscopy can help, but systematic discrepancies hinder model interpretation. In this study, we evaluate the performance of semi-analytical line transfer (SALT) and empirical partial covering models (PCMs) to recover the properties of outflows in the FIRE-2 simulation suite from synthetic Si II lines (1190 $\AA$, 1193 $\AA$, 1260 $\AA$, 1304 $\AA$, 1527 $\AA$). When applicable, we assess each model's ability to recover mass, energy, and momentum outflow rates, as well as radial density and velocity profiles, column densities, and flow geometries. We find that the PCM underestimates column densities by 1.3 dex on average in the range $15 < \log N\ [\text{cm}^{-2}] < 17$ with dispersion 1.3 dex. We attribute this bias to instrumental smoothing. Since the PCM underestimates column densities, it also underestimates flow rates, though its predictions are independent of radius, with a dispersion of 0.55 dex. We detect no bias in the SALT estimates of the column density with dispersion 1.3 dex. When the velocity and density field obey power laws, SALT can constrain the mass, momentum, and energy outflow rates to 0.36 (0.63), 0.56 (0.56), and 0.97 (0.80) dex at $0.15(0.30)R_{\text{vir}}$, respectively. However, certain profiles in FIRE-2 fall outside the SALT framework, where the model breaks down. We find that SALT effectively tracks the flow geometry, capturing the temporal evolution of the photon escape fraction that is out of phase with the star formation rate, fully consistent with hydrodynamic simulations. We advocate for integral field unit spectroscopy to better constrain flow properties.