This paper presents analytical results on power profile estimation (PPE) methods, which visualize a signal power evolution in the fiber-longitudinal direction at a coherent receiver. Two types of PPE methods are reviewed and analyzed, including correlation-based methods (CMs) and minimum-mean-square-error-based methods (MMSEs). The analytical expressions for their output power profiles and spatial resolution are provided, and thus the theoretical performance limits of the two PPE methods and their differences are clarified. The derived equations indicate that the estimated power profiles of CMs can be understood as the convolution of a true power profile and a smoothing function. Thus, the spatial resolution and measurement accuracy of CMs are limited, even under noiseless and distortionless conditions. Based on this fact, closed-form formulas for the spatial resolution of CMs are presented. On the other hand, in MMSEs, such a convolution effect is canceled out and thus the estimated power profiles approach a true power profile under a fine spatial step size.