Thank you @zurvanic and @andrarchy for the update... I'll echo what a few others have pointed to... it's nice to see an announcement with a little meat on its bones!
Marketing-- not entirely sure about the viability of crowdsourced marketing, but I'd submit that we already HAVE a "crowd" of pretty dedicated users (quite a few of whom are also considerable influencers in their niches) who are just waiting for a good reason to start telling the world about Steemit.
What's that good reason? Well, there has to be "something here" that's worth telling the world about-- and that means some better "social tools" if we are truly to present Steemit as a viable player in the social space. Communities are a great start there. I am also very hopeful that you mention a referral program... I just wrote a "suggestion box" post about this, a couple of days ago.
The TL;DR version: Let's work on the "gamification" of onboarding new members to the community; using not a monetary (Steem) reward so much as a "status" reward of being a top referrer; second, let's reward referrers to bring ACTIVE users, not just "new accounts." End TL;DR
Marketing II: Let us not be too afraid of "organic" growth. Doing a big splashy campaign can backfire in very unfortunate ways... I've seen far too many user generate content sites crash and burn by trying to grow too fast. Above all, step with EXTREME CARE when using the "blogging for rewards" idea... the "clicking buttons for cash" crowd can descend like a swarm of locusts and turn a venue into a cesspool of garbage in a matter of 60 days... from which there is NO COMING BACK. Better to build a quality base, more slowly. Also, in bringing people onboard, emphasize Steem as a a path to becoming a stake holder (long term engagement) rather than a cash reward for tomorrow's pizza (short term engagement).
The Steem Ecosystem of Apps: This sounds promising... the one thing I would STRONGLY stress is the importance of everything launching/running inside the "flagship app," namely Steemit. It's nice that we're all decentralized and open source... but in order to gain broad-based support and acceptance, there needs to be a "central" Steemit interface from where everything runs.
Think about it: Just like people are "on Facebook," they need to be able to parse this user experience as being "on Steemit," and NOT "being on 47 individual apps." That kind of fragmentation will just ot fly with anything beyond a hardcore niche of developers and blockchainiacs.
Building Communities: I know this has been said 1000 times, but I'll say it again: Communities are NOT built by code, they are built by PEOPLE. Significance? As you go forth, keep the major content creators in the development loop, not just the developers. The people who "do social" for a living, but might not know code from an elephant's butt.
I sincerely hope "Communities" will be incorporated as "channels" of sorts... remember, lots of people have multiple interests.
As for the Reddit reference, I guess Reddit is testament to the fact that you can build a large follower base with an outdated Web 2.0 interface that never gets updated. That said-- and Reddit's high rank with Alexa notwithstanding-- Steemit is SO much more than a Reddit can ever hope to be. Above all (and someone else brought this up in comments) let us remember that Reddit is rather USA-centric and white-male-under-30-centric, and we have the potential to reach a truly GLOBAL market with Steemit. Just look at the explosive number Aceh posts, happening more or less spontaneously. Point? International Communities, as a subset of "Communities."
Let's not copy Reddit too much... Reddit is FAR better at "being Reddit" than we can EVER hope to be, so let's focus on being Steemit!
Anyway, very happy to see this "in the works" document come out! Feels like something is actually happening, after what feels like a long slumber!
Great points, thanks for sharing this detailed analysis! Agree with most of your comments. Steem on!