Planetary remote sensing of regolith surfaces requires use of theoretical models for interpretation of constituent grain physical properties. In this work, we review and critically evaluate past efforts to strengthen numerical radiative transfer (RT) models with comparison to a trusted set of nadir incidence laboratory quartz emissivity spectra. By first establishing a baseline statistical metric to rate successful model-laboratory emissivity spectral fits, we assess the efficacy of hybrid computational solutions (Mie theory + numerically exact RT algorithm) to calculate theoretical emissivity values for micron-sized α-quartz particles in the thermal infrared (2000--200 cm-1) wave number range. We show that Mie theory, a widely used but poor approximation to irregular grain shape, fails to produce the single scattering albedo and asymmetry parameter needed to arrive at the desired laboratory emissivity values. Through simple numerical experiments, we show that corrections to single scattering albedo and asymmetry parameter values generated via Mie theory become more necessary with increasing grain size. We directly compare the performance of diffraction subtraction and static structure factor corrections to the single scattering albedo, asymmetry parameter, and emissivity for dense packing of grains. Through these sensitivity studies, we provide evidence that, assuming RT methods work well given sufficiently well-quantified inputs, assumptions about the scatterer itself constitute the most crucial aspect of modeling emissivity values.