As to the extent of rewards inuring to a handful of accounts, this is a chart provided by @aggroed just before HF19.

The figures I cited came from the post where I got the chart, IIRC. Regardless of the mathematical precision of the stated figure, the chart very clearly shows that those figures cannot be far off.

As I recall, @aggroed estimated that after HF19 he expected author rewards to improve such that 93% of author rewards would be captured by less than 10% of accounts.

I have not addressed bots, and do not know how they may be, or not be, contributing to this concentration of rewards in so few accounts. However, HF19 reduced the number of votes requisite to fully draining vote power to 10 from 40, which to me indicates that reducing the manual voting necessary to maximize financial rewards of self voting was the intention, as bots don't care how long it takes.

This would also reduce the necessary number of posts required to make, in order to have posts to self vote, and those posts are not written by bots.

My vote, as you can see, is insubstantial as a means of discouraging the production of content. I do not agree that it represents a meaningful force in comparison to votes backed by $1000's, or $M's of SP.

Thanks again for helping me to better understand these issues.