Taken by itself, that observation ("this phenomenon arises from individual behaviour rather than collectively emerging due to pedestrian-pedestrian or pedestrian-boundary interactions") doesn't quite rule out the hypothesis. I'd need to know more.
For example, I don't know how big the "circles" are. If the pedestrians are found to walk in relatively small circles, then they are effectively "patrolling" a small area of space and they know that any new pedestrians (or other threats) are necessarily going to emerge from outside of that area; they're going to keep their attention oriented primarily outside the circle and are going to keep their right side facing outside. So the very first pedestrian to enter the space will instinctively start circling counter-clockwise, even if there are no other pedestrians to influence his behavior. On the other hand, if the "circling" behavior consists of walking the perimeter of a large room, then you'd expect the pedestrian to circle clockwise, and the "self-defense" idea wouldn't fit.
Again, I don't know if the data supports, contradicts, or is neutral to the "dominant side is for self-defense" idea, I'd need to read the actual article in depth (sadly I don't have time to do that this morning). So it's just a bit of speculation on my part.