I believe in the potential of Steemit but....
The problem for Bitcoin, then Ethereum and now Steemit is that the ideas behind these things are revolutionary and highly interesting but also highly complex and confusing. The average person doesn't have time to understand the in's & out's of the technology, so it's simply ignored.
Steemit has been called 'the killer app' over and over and for good reason, the concept of being paid for content creation is amazing and is sure to draw attention. But once attention is drawn it quickly becomes clear that although you can earn a number on a screen for content creation actually being able to buy tomorrows lunch with today's blog post requires a knowledge of cyrpto and exchanges that many laymen's simply don't have time for/don't want to learn.
How to turn 29,000 accounts in 2.9 million accounts and beyond
Cash, money. The ability for the average user to turn Steem or Steem dollars directly to cash would make Steem a true 'killer app' and would see mass adoption quicker then anyone could keep up. Obviously within crypto one of the biggest issues is turning 'coins' into cash, there are so many regulatory barriers the process can be painful at best, especially for those outside the US.
This doesn't mean this is impossible. Steemit is also lucky in its positioning within crypto as its hardly seen as one, this means Steemit may have more of a chance of teaming up with money services like Paypal or Skrill to create what would easily be the biggest 'app' of 2016.
The Steemit Team and what they can do to encourage adoption
The Steemit team pre-mined a huge amount of Steem (which although turned me away from Steem originally now looks like an opportunity to use these funds to slingshot adoption) and most defiantly have the funds they need to advertise on a mass scale, create fantastic API's and encourage exchanges to be cooperative.
Steemit has the most potential in all crypto space at the moment but the average user doesn't know how to turn their Steem > Bitcoin > Spendable Money, if this problem can be solved I'd be willing to bet we'd see unheard of levels of adoption globally.