It's been over three months since I wrote up a personal list of things I'd like to see Steem improve. There were some fairly drastic changes there, a lot of which would take years to implement, if ever. So, here's a couple of things that could be addressed in the short to medium term. Over the last 3-4 weeks I have been active on Steem again, blogging and curating, which has given me a more accurate perspective. Part of the curating process has been to go through every single post made on Steem, but later aided by Steemlookup. It's a must-use app for any curator on Steem.
Back to Steem, today... it's a complete shitshow. There's very, very little content, almost no engagement etc. The only positive is that there's so little content a lot of it is being found and rewarded. In a way, this reminds me of late-2016 to early-2017. Oh, and the price is similar too, ouch. There's a bit more usage, but unfortunately, there's also greater toxicity, so it's not all an improvement since then. Still, it's not quite as bad as mid-2016 when only a few whales' pets took home 90% of the rewards, or early-2018 when bidbots were profitable and dominated all, but it's still far and away the most terrible social media experience on the internet. The other kinda positive is that there are definitely some niche communities that have remained, so there's a very, very small but engaged and loyal userbase. Steem's best days I'd say were Q2 2017. There was a growing userbase and optimism, the whales had agreed to sit out voting so we saw by far the most effective curation in Steem's history. As with all good things, that ended, and since mid-2017, Steem has been in a downward spiral as a social network with only minor improvements.
I'll leave the same qualifier as my other post - this is purely a personal view. I don't have the answers, and everything I say might be totally wrong. Please consider it a series of random thoughts. So with that, let's get started.
The first thing we will assume in the short-term or medium-term is not canceling the reward pool outright. While I still believe this concept has failed, we'll assume for the time being that the common reward pool will continue to be around, or at least be a "default SMT" that'll work similarly to how it does today. Also, there are no minor adjustments like considering in things like changing the ratio of curation to author reward, reducing reverse auction window etc; instead looking at the larger picture. It has been obvious for years now that while those adjustments may have benefits, the real problems of Steem lie with the fundamentals.
Reduce inflation to 2%-5%.
This one is pretty simple. STEEM's inflation is too high and has led to a perpetual state of price declining save for the very short-lived bull markets, and other short-term impulses (like the Tron buyout rumours recently). Cut inflation across the board. The SP interest can be eliminated entirely, reward pool will see the greatest cuts etc. While this will be undoubtedly unpopular in the short term, it's necessary for long-term sustainability. Witnesses and content creators have seen dramatic reductions in the past, not to mention a 99% decrease in price.
Reform SBD.
SBD has been a waste thus far, but 
Firstly, for author reward SBD printing, following the 3-day moving average has to go, replaced by a very conservative, 2-4-year moving low average. Mind you, the witness price feeds and 3-day averages still stays for other purposes, but from the historical price feed, a new metric should be used. I don't have an algorithm to offer, but let me explain some general ideas. The problem with altcoin markets is that it's typically 2-3 years of very low prices followed by 3-6 months of insane overvaluation. This base reality needs to be factored in to avoid printing gazillions of SBDs and taking on an unreasonable debt during these short time periods that'll remain a burden for several years to come. Instead, even if the price were to pump to $10 in the next 3 months, the SBD printed would consider a long-term low-average price of $0.20 or whatever. This can go up very gradually over the years, instead of simply going with what the 3-day average is. The surplus can be printed in SP, or simply temporarily reduced SBD printing which will further help with reducing inflation.
Build a DeFi platform around SBD with STEEM as primary collateral. This is a mechanism that wasn't around when SBD was first considered. But today, it's something that has proven to work. Let the market decide interest rates and work to stabilize SBD through borrowing and lending. There's some open source code out there, so I'd imagine this is achievable medium-term. The major drawback, of course, is anyone even going to trust a STEEM-backed DeFi system given how terribly this economy has failed? A few fanboys will, and maybe that's enough to kickstart it on a path towards regaining some confidence. Of course, conversions (now bi-directional), interest rates and all of those mechanisms will now be dictated by the DeFi platform.
Bonus: Reduce SP period
Yeah, we have talked about this enough. I have yet to see a rational argument against it, but it's nevertheless controversial. Perhaps starting off with a 4 week power-down period could be a good compromise.
Bonus: Privacy
It has been obvious for a long time, but there's particularly been a lot more talk about privacy in the last 3 months since my initial post. Steem absolutely must build a platform with privacy controls. Really, this should be the #1 priority, but I'm putting this as a bonus because I don't have the engineering perspective, so I don't know what's required to accomplish this. Alternatively, Steem can remain a zero-privacy solution, but that'll only ever attract a very, very niche audience.
With Twitter financing a team to create an open standard for social media, the clock is ticking. No one's going to adopt a niche quasi-centralized blockchain solution if there's an open standard that'll be used by billions. Not to mention, Voice beta is coming soon. (I have very little expectations for Voice, but it'll surpass Steem overnight due to EOS' network effect and Block.one's formidable marketing prowess. I bet the marketing budget for Voice is more than Steem's entire market cap, and even that's but loose change for them.)
Longer term, even the above changes are thoroughly unsustainable, but we'll cross that bridge some other time. Right now, any progress is progress.
On a positive note, I'm glad to see Communities is finally shaping up. I've been asking for it since 2016, shouting about it all through 2017, look forward to see it in production some time in 2020.