Frequency-domain data continues to be widely employed in modal analysis, primarily due to its simplicity and intuitive interpretation. In particular, for simpler systems, an experienced practicioner can often directly discern system modes from frequency response functions. However, widely used methods such as rational fraction polynomial (RFP) and least-squares complex exponential (LSCE) can suffer from ill-conditioning, potentially obscuring critical structural dynamics. Meanwhile, time-domain techniques like subspace state-space system identification (N4SID) become expensive when analyzing large systems, raising concerns in aerospace applications demanding high accuracy and robustness. As a solution, the Loewner Framework (LF) has emerged as a robust, highly noise-robust alternative for extracting modal parameters. Initially, LF focused on single-input multi-output (SIMO) system identification, particularly in applications related to structural health monitoring. Originating from research on model order reduction in electronics and aerodynamics, the efficiency of the Loewner Framework (LF) has been effectively demonstrated across aeronautical, mechanical, and civil engineering applications. Furthermore, advanced formulations of LF now support the analysis of multiple-input, multiple-output (MIMO) and output-only datasets, showcasing its scalability. This work revisits the theoretical foundations of LF, critically assesses its advantages and limitations, and proposes avenues toward achieving comprehensive, efficient, and automated modal analysis.