Mass spectrometry-based proteomics has advanced with the integration of experimental and predicted spectral libraries, which have significantly improved peptide identification in complex search spaces. However, challenges persist in distinguishing some peptides with close retention times and nearly identical fragmentation patterns. In this study, we conducted a theoretical assessment to quantify the prevalence of indistinguishable peptides within the human canonical proteome and immunopeptidome using state-of-the-art retention time and spectrum prediction models. By quantifying the proportion of peptides posing challenges to unequivocal identification, we set the accessible portion within a given proteome, thus establishing the current boundaries of the field, and underscore the effectiveness of contemporary analytical methodologies in resolving the complexity of the human proteome and immunopeptidome via mass spectrometry.