The rate and mechanism of mass loss of red supergiants (RSGs) remain poorly understood. Motivated by a new empirical mass-loss rate prescription based on the largest and most complete sample of RSGs in the Small Magellanic Cloud (SMC), we investigate the impact of various established and recent prescriptions that span a wide range of RSG mass-loss rates on the evolution, the observable properties, and the final outcome of single massive stars. Our results show that higher mass-loss rates result in earlier envelope stripping and shorter RSG lifetimes, particularly for the more luminous stars, leading to a steeper luminosity function and predicting hotter final positions for the supernova (SN) progenitors. None of the considered mass-loss prescriptions is fully consistent with all observational constraints, highlighting persistent uncertainties in deriving and modeling the mass loss of RSGs. Models incorporating new theoretically motivated high mass-loss rates predict overly efficient envelope stripping, conflicting with the observed populations of luminous RSGs and detected SN progenitors, while commonly used moderate and weak rates tend to overestimate the number of luminous RSGs. The enhanced mass-loss rates for luminous RSGs predicted by recent multi-metallicity empirical studies improve the agreement with the observed RSG luminosity function in the SMC and naturally reproduce the updated Humphreys-Davidson limit. However, they also produce an excess of luminous yellow supergiant progenitors, which remain undetected in nearby SNe. We also estimate that binary interactions tend to slightly enhance the formation of luminous RSGs due to mass accretion or merging. Our study examines the impact of RSG mass loss during the late stages of massive stars, and we highlight the significance of using comprehensive observational data, exploring the uncertainties involved, and considering the effects of binary-induced or episodic mass loss.