Rest-frame polarisation measurements of thermal TDEs (Patra et al. 2022; Leloudas et al. 2022), BFs, and the extreme coronal line emitter AT2022fpx (Koljonen et al. 2024). Time is scaled by each event’s rise time to peak bolometric luminosity to remove the primary timing dependencies on SMBH and stellar mass (i.e. $t_{\mathrm{peak}}\propto M_{\mathrm{BH}}^{1/2}{M}_{★}^{-1}{R}_{★}^{3/2}$). Here, the t_Rmax is the epoch at which the MOSFiT-derived blackbody radius (and, to a good approximation, the bolometric luminosity) reaches its maximum. Polarisation measurements at Δt/t_peak < 0 correspond to the optically thick expansion regime of the electron-scattering photosphere, where polarisation reflects the asymmetry of the outermost layers. The dotted line at Δt/t_peak = 0 marks the photospheric turnover at Δt/t_peak > 0, where the photosphere recedes as the ejecta become increasingly more optically thin. This allows asymmetries in the inner ejecta to become more apparent in their polarisation imprint. Note that AT2019qiz is corrected by an estimation of host depolarisation using ZTF data and assuming P_host ≈ 0%. Within our sample, measurements with P_obs/σ_P < 3 are reported as 2σ upper limits (empty markers).