I will present recent results from the IRAM 30m large program LEGO, studying star-forming regions in the Milky Way across a wide range of environments to help interpret extragalactic observations of molecular line emission. Our recent work has focused in particular on several prominent, massive star forming regions, serving as excellent templates of regions bright enough to be observed in other galaxies. I will show new results probing canonical extragalactic molecular gas tracers like e.g. N2H+, HCN, HCO+, HNC or CS and their emissivities in particular as a function of gas density. Of particular interest is the scale dependence of such lines and canonical ratios from sub-cloud to cloud scales, where we find HCN (also HCO+, HNC, CS) having a roughly fixed ratio with N2H+, the Galactic "gold standard" tracing star forming gas, motivating the use of observationally cheaper HCN or HCO+ observations as proxies for such denser molecular gas at scales larger than that of individual clouds. Similarly, and comparing to recent extragalactic surveys of the same lines, canonical line ratios like HCN-to-CO are found to scale consistently with other gas density metrics from parsec to kiloparsec scales. These and related results highlight the potential of finally bridging the gap between star forming regions in the Milky Way and other local galaxies, by combining surveys like LEGO with recent, high-resolution surveys in local galaxies using a matched set of observational tracers.