Modeling of the spectral line energy distribution (SLED) of the CO molecule can reveal the physical conditions (temperature, density) of molecular gas in Galactic clouds and other galaxies. Recently, the Herschel Space Observatory and ALMA have offered, for the first time, a comprehensive view of the rotational J = 4-3 through J = 13-12 lines, which arise from a complex, diverse range of physical conditions that must be simplified to one, two, or three components when modeled. Here we investigate the recoverability of physical conditions from SLEDs produced by galaxy evolution simulations containing a large dynamical range in physical properties. These simulated SLEDs were generally fit well by one component of gas whose properties largely resemble or slightly underestimate the luminosity-weighted properties of the simulations when clumping due to non-thermal velocity dispersion is taken into account. If only modeling the first three rotational lines, the median values of the marginalized parameter distributions better represent the luminosity-weighted properties of the simulations, but the uncertainties in the fitted parameters are nearly an order of magnitude, compared to approximately 0.2 dex in the best-case scenario of a fully sampled SLED through J = 10-9. This study demonstrates that while common CO SLED modeling techniques cannot reveal the underlying complexities of the molecular gas, they can distinguish bulk luminosity-weighted properties that vary with star formation surface densities and galaxy evolution, if a sufficient number of lines are detected and modeled.