Déjà voodoo or new trails ahead? Re-evaluating the mixing typology model