Figures illustrating the observational limitations of our sample. GOLD, SILVER, and BRONZE objects are represented by triangles, crosses, and squares symbols, respectively. Left: Lyα luminosity as a function of redshift for the parent sample. The horizontal black dash-dotted line represents the luminosity cut (L_Lyα = 3 × 10⁴⁰ [erg s⁻¹]) used to mitigate the redshift dependence. Middle: B/T flux ratio as a function of S/N (calculated as described in Sect. 3.5) for the whole double-peak sample. B/T ≈ 0 objects can be seen in Appendix D. They have a very small blue peak. The horizontal grey dashed line at B/T = 0.5 marks the blue-dominated versus red-dominated dividing line. The vertical black dash-dotted line represents the S/N cut (S/N = 7) made to balance the demands of being able to detect extreme B/T and keeping a representative subsample of double-peaks. The dashed lines mark boundaries of 1/(S/N) < B/T < 1–1/(S/N). Three objects have a B/T > 1 because the total flux is smaller than the flux in the blue peak because the continuum is negative between the peaks (see Appendix D). Right: Peak separation as a function of redshift for the whole double-peak sample. The horizontal black dash-dotted line represents the v_sep cut (v_sep = 150 km s⁻¹) used to account for the spectral resolution of MUSE. For both the Lyα luminosity and the v_sep, the observational sensitivity and spectral resolution curves are not straight lines. The ones shown in the plots are our selection limits.