[aerohead]

Perhaps an unfortunate use of the word.
* A skyscraper would never experience a sub-critical Reynolds number, and a laminar boundary layer, so introducing the notion of a golf ball dimple parallel would be inappropriate.
* Every 'face' of the tower would have to be considered within the context of the particular type of flow it experiences.
* With increasing height, the moment arm is lengthening, and the 'narrowing' reduces the local frontal area proportionately, mitigating the torque introduced by the aerodynamic drag from wind speeds which are no doubt, also increasing with elevation.
* At times, edge vortices might induce torsional loads.
* Natural frequencies of the structure would vary with location. These frequencies, and their harmonics would need to be dampened or eliminated, in order to avoid flutter.
* The staggered 'stepping' would work towards that end.
* The edge flow separation may not reattach at all near the top of the tower at times, imparting a very low base pressure in the turbulent wake, and amplifying the overturning torque; putting part of the structure and foundation in compression on the leeward side, while inducing an opposing tension on the windward face.
* The staggering of the steps 'flow confusion' might be an analogue to tuning downwash and up-wash on the rear of a car in order to 'burst', otherwise strong longitudinally-attached vortices.
* I'll bet that the load cell data is invaluable.