As I said before, Steemit is an early stage experiment which could eventually fail in any number of ways. Thus far however, evidence suggests it is gaining an enormous momentum. Not only are the current data points impressive, but when you understand that all these results have been achieved in roughly 4 months from launch, you can only imagine what can happen in one, two or three years.
Steem solves so many problems that have long been subjects of discussion on discussion forums since the inception of discussion groups, problems that have grown in scale the more people got involved. The adaptations to the algorithms that help moderate the discussions can be applied quickly and rapidly accelerate the process. This is why Steemit is growing so fast. It has already changed several times with new improvements.