We employ a Monte Carlo radiation transfer code to investigate the multiwavelength photopolarimetric variability arising from a spotted T Tauri star surrounded by a dusty circumstellar disk. Our aim is to assess the ability of the magnetic accretion model to explain the observed photopolarimetric variability of classical T Tauri stars and to identify potentially useful observational diagnostics of T Tauri star/disk/spot parameters. We model a range of spot sizes, spot latitudes, inner disk truncation radii, and system inclination angles, as well as multiple disk and spot geometries. We find that the amplitude, morphology, and wavelength-dependence of the photopolarimetric variability predicted by our models are generally consistent with existing observations; a flared disk geometry is required to reproduce the largest observed polarization levels and variations. Our models can further explain stochastic polarimetric variability if unsteady accretion is invoked, in which case irregular (but correlated) photometric variability is predicted, in agreement with observations. We find that variability in percent polarization is by itself an unreliable diagnostic, because certain system geometries do not produce any variability in linear polarization (contrary to the commonly held notion that hot spots will necessarily produce periodic polarimetric variability). Observations of variability in polarization position angle, however, could provide useful constraints on system inclination. The observation of wavelength-dependent polarization position angles, attributed by some to interstellar effects, is naturally explained by our models. Certain system geometries yield peculiar photometric light curve morphologies, the observation of which could also serve to constrain system inclination. We do not find useful diagnostics of disk truncation radius, nor do we find significant differences when we model the spots as rings. We also investigate the reliability of modeling spot parameters via analytic fits to multiband photometric variations. We find that commonly used analytic models consistently recover input model parameters but that inferred spot temperatures are more sensitive to uncertainties in the photometric data than previous modeling would suggest.