I’ve found the error in theorem 16.216 of my book. The assumption $(\mathsf{RLD})_{\mathrm{in}}\uparrow^{\mathsf{FCD}} Y = \uparrow^{\mathsf{RLD}} Y$ was wrong.