Roskam describes an approach for optimizing the sweep angle of a vertical tail with fixed aspect ratio and mounting point. Basically, increasing the sweep decreased dCLa, and increases moment arm; increase sweep until these factors are balanced.
His accompanying image, though, shows this at a reasonable cruise angle of attack. Shouldn't this be performed at the most critical flight condition for tail authority, and isn't this usually Vs0 for for a single engine? If so, this leads to a pretty substantially different result, as much as 10° different.