One of the largest surprises from the LIGO results regarding the first gravitational wave detection (GW 150914) was the fact the black holes (BHs) were heavy, of order 30 - 40 solar masses. The most promising explanation for this obesity is that the BH-BH merger occurred at low metallicity (Z): when the iron (Fe) contents is lower this is expected to result in weaker mass loss during the Wolf-Rayet (WR) phase. We therefore critically evaluate the claims for the reasons of heavy BHs as a function of Z in the literature. Furthermore, weaker stellar winds might lead to more rapid stellar rotation, allowing WR and BH progenitor evolution in a chemically homogeneous manner. However, there is as yet no empirical evidence for more rapid rotation amongst WR stars in the low Z environment of the Magellanic Clouds. Due to the intrinsic challenge of determining WR rotation rates from emission lines, the most promising avenue to constrain rotation-rate distributions amongst various WR subgroups is through the utilisation of their emission lines in polarised light. We thus provide an overview of linear spectro-polarimetry observations of both single and binary WRs in the Galaxy, as well as the Large and Small Magellanic Clouds, at 50% and 20% of solar Z, respectively. Initial results suggest that the route of chemically homogeneous evolution (CHE) through stellar rotation is challenging, whilst the alternative of a post-LBV or common envelope evolution is more likely.