Could be construction....could be other things. The structural design of these towers yields their circularity of the main support. If that is compromised by incomplete connection of the sections OR even just a variation in the wall thickness just sets it up for eventual failure.
Notice how clean the failure is at the bottom flange... it’s as if all the bolts instantly disappeared. There is no distortion on that flange, no ‘tears in the metal’, nothing. This is a dead giveaway that the bolts all failed in cyclical fatigue. What is amazing about this failure is that it occurred after only four years and it came down in what were reportedly windless conditions. This essentially indicates that the bolts had already ‘almost failed’ failed to the extent that even a minor change that affected the balance was enough to bring it down. Once the first bolt went, it would have been like sitting beside a popcorn maker as the other ones quickly popped.
Were there other factors? Four years is such a short period of time that one wonders if the right bolts were used or if a proper bolt tightening technique was used. Not paying attention to those issues will greatly hasten how fast cyclical fatigue can occur.