Going through the cited sources, I'm more convinced the defender's advantage theory is right than before I started. First the claim that stairs would not be op

Going through the cited sources, I'm more convinced the defender's advantage theory is right than before I started. First the claim that stairs would not be optimized for combat because fighting on stairs was undesirable holds no weight. Castles were designed with many layers of defense - it was entirely expected that large sections of a castle would be lost and the defenders would continue to hold out in other sections, bitterly holding chokepoints. From murder holes to machicolations, every inch of castles were optimized defense. Even if in practice defenders would prefer not to fight on a stairway, if all else were equal (and it's hard to think of something more arbitrary than which way the stones are flipped), why wouldn't they go with the option that potentially could be helpful instead of potentially aiding an attacker? Next, the argument that "not all castles had clockwise staircases" seems to be an own-goal. During the time periods when having to defend a castle was likely, staircases were overwhelmingly clockwise. It is only in the late middle ages when defense became much less of a priority that anticlockwise stair cases start gaining popularity, and the later it gets the more common anticlockwise stairs become. If there were some non-military utilitarian reason for the choice, such as making it easier for someone to steady themselves or carry lanterns, presumably that need would remain. If the choice were non-utilitarian from the beginning, why the initial disparity? No doubt the designers of these buildings had multiple competing concerns, including aesthetics and convenience, but clearly the balance shifted. Examples of anticlockwise stairways were common in structures not intended for defense in earlier structures, which means the shift was not technological and further makes non-military utilitarian requirements unlikely. Obviously it would be nice if we had more surviving sources from the time period, but it's hard to imagine any other theory fitting the data so well.