I think Steemit has certain characteristics that are right on the money. But if it wants to succeed, we must be honest: There are other characteristics that need fixing.What follows might be a bit "harsh" as a criticism, but it is only to wake people up from the confirmation bias of being in the upper technological tier and thinking that some things are ...normal (when they aren't).I will be addressing these points below - with the main point being a necessary reduction in complexity:

Three currencies: JUST NO. I can’t emphasize this enough. The currency should be one.

The complexity of the design should not be of interest to the user. The user does not want to read a 40+ page whitepaper or hear why he must juggle 3 currencies in a "for noobz" guide - which really understands that this is absurd but tries to convince the user that it isn't. The complexity should not be a concern for the user. Dealing with complexity should be a ‘background’ process while the user only has to face a ‘foreground’ interface that is easy and intuitive. For example three currencies mentioned above. Make it one currency and just use a button like “Steems locked” or something for “locking” them – and even there, have a (i)nformational mouse-over that the user can click to learn why it’s good to do so and what this will do. The Permissions screen is equally problematic. No explanations whatsoever, unintuitive functionality even for cryptocurrency experts (who are already familiarized with complexity). I understand the system is “beta” but this is not good in terms of user interface. I can't even find what some things do, even after googling, reading the whitepaper, or reading bitcointalk.
The system as it is right now has a heavy confirmation bias from the crypto-world complexity. Devs, being cryptocurrency-experts, apparently think that this kind of complexity is “normal” for everyday use. It isn’t. The system is not ready for “prime-time”. If the system is targeting just cryptoguys, well, ok, it might capture some small segment of the market, or work for pump & dump purposes, but is this what we are here for? Social platforms are successful because they are extremely easy to use. Even the "necessity" of having to read 5-10-20 minutes of theory before using a platform is an excessive demand. Let alone if the content is complex in itself.
Why is there no proper content editor where the user can use bold, italic etc? I clicked on something it said it supported Markdown and then it said it is...Github syntax. Frickin' Github? Seriously? I tried to move a document from word to the platform, it kept no bold/italics etc… it was like pasting from notepad. Should I be an “expert” to post a ...nice looking text? Why? I don’t want explanations like “but, but, if you just do two ** you’ll get bold”. Yes, and if I write [b] in a forum I get bold, but I just click a button to bold stuff. It’s the way it should be done.
The user economy (poster-reader-voter etc), per the whitepaper which says that advertising is replaced by users etc etc, is a closed loop. Targeted advertising revenue could be used to tap into external revenue streams. Viewing ads should be optional (per user preferences). Those who want to view ads might get paid more or good content creators might get a chunk out of the ad revenue generated by their posts viewership.
Online guides, help buttons, “what’s this” buttons etc are needed to deal with current complexity levels but a lot of this should ultimately be changed. Anything that your mother can’t figure out should be taken to “background complexity” with a nice easy front-ending simplicity.
Inability or lack of flexibility to cure problems will lead to replacement. The above suggestions are critical for success. If the design is problematic, another one will come along, clone the design, fix the problems, make it more profitable by integrating external revenue streams and then say “See? We are like Steemit but unlike them we are better in these areas – so now we are 10 times more usable and more attractive to users, ensuring that we will be way more successful.” At that point Steemit is DOA because it had a good idea that it did not manage to convert to a practical design.

The above is just my brutally honest criticism to make the idea work properly. Patting on the back is appropriate for the idea of rewarding content creators and merging cryptocurrency but criticism is due if this is not done in a simple, efficient and seamless way.