RE: The Growing Steem Middle Class
There is an inherent question which this analysis does not and perhaps cannot reveal, and that's how much of the "lower classes" beneath whales are real, active accounts driven by people and which ones are the inheritors of delegation provided by a larger account, but essentially a puppet.
Because that makes a difference.
It is certainly feasible for the upper classes to appear to diminish in terms of population while effectively simply being a larger percentage of the population, via the mediation of bots and other stake holding non-human accounts. In some cases, this is quite useful, as when non-human agents act as community hubs and community actors, driven by external funding. In some other cases, in particular anti-content bot swarms which act to vote down particular users or content, this is less than desirable.
I accept that it may be impossible to determine whether any given account is one of these classes and thus impossible to determine whether the migration of wealth into the mid range is the result of the allocation of the reward pool from below or the delegation of wealth from above.
But this is definitely a question that needs to be taken seriously.
You're right. I use the word "account" where possible rather than users, because it is distribution across accounts which a single user can have many of (I have several, and it is the norm especially for whale users).
There are some methods that can be used to identify when a user is not the same as another, or not a bot etc. but none that work on this scale of analysis. SMT's will be bringing more identity-based features which will help.
I'm not sure that SMT's will help with the identity problem in the long term since they will ultimately make it even easier to distribute "identity" across multiple communities, some of which will not necessarily intersect. We end up with the same problem that exists now, except potentially even harder to chase down.
But in light of the identity problem, it may be difficult to actually speak reasonably about the percentage of upper echelon accounts on the blockchain. We need to know about the ratio of increase of actual tokens over that time (which is, at least, a fixed amount thankfully), compared to the distribution dynamically, with delegation taken into account.
While financial analysis isn't my field, recognizing the overall problem suggests that any analysis which tries to leverage identity is pretty much doomed to irrelevance. While we might like to take heart from the news and suggest that it means there is a rising middle class – we can't.
At least not as shown by this data.