I think this is the case. A siege is just one scenario out of many where someone could be attacking you in a castle. I have also been in several castles in UK where the penultimate step before the roof is deliberately 1 inch higher than all the others. The story is that if you were being chased up the stairs your attacker would trip over onto the roof and you could easily dispatch him. Would this author also claim the rationale of this feature as a myth, and that they just made the same measurement mistake in multiple castles?

Incidentally, if you're not aware of it, a 1 inch difference in stair height will almost certainly make you trip up. I did myself when I was climbing them at first.

Also the notion that once attackers were in a castle it's all over is not true. Late medieval castles were designed to be modular, layered and flexibly defensible. The enemy might get into one part, but you could retreat to another part and defend that. Might give you enough time to be relieved by an external force. Have a look at Caerphilly castle as an example: https://en.wikipedia.org/wiki/Caerphilly_Castle

Finally, it seems that no real explanation is given to the bias towards building clockwise staircases. Why would this be the case? Doesn't it just mean that wherever possible they would build the staircase clockwise because there was some inherent advantage?