Welcome to Steemit. I can certainly appreciate your feelings and concerns. I can just briefly give my perspectives, since I've been around for a while (since before the first payouts). The very first thing to mention is that these types of systems are difficult to design and predict. Especially since there are potential financial rewards on this platform, it essentially becomes one large, complex problem of game theory, which is largely complicated by the large number of participants. Suppose you want to design how Steemit works, and so you try to design the platform to incentivize "good" behavior, such as upvoting "good" content and not just blindly upvoting everything. So you decide to give curators a percentage of a posts total payouts. But then people just start voting for things that already have high pending payouts, so they can get a percentage of the payout. So then you decide to reward the earlier voters more than the late voters, to incentivize IDENTIFYING "good" content, rather than just VOTING for "good" content. But then people just start using bots to vote as early as possible for every single post by popular authors, since they expect those to become valuable, creating a huge advantage for bots which basically guarantees popular writers high payouts, and greatly decreases the chances other authors will be identified. So then you change the rules to a kind of reverse auction style voting, where early voters still receive a higher percentage, but within the first 30 minutes of voting a certain percentage of the rewards you would get go instead to the author, and that percentage decreases linearly from 100% to 0% over 30 minutes.

This hypothetical reasonably describes the actual process that has occurred over the (brief) development of Steemit. Although the system is far from perfect, it continues to be upgraded, and new changes are experimented with. We have already undergone 19 hard forks of the code to try to better incentivize the platform and fix certain issue. For example, you were mentioning that there is a huge disparity between the influence of whales and minnows, but that disparity was exponentially larger before the previous hard fork. Now influence is exactly proportional to Steem Power, which can hardly be considered unfair. Would you expect someone with a brand new account who signed up from Facebook to have the same influence as someone who purchased $500,000 worth of Steem Power? Furthermore, whales have far more to lose or gain from the appreciation of the Steem tokens, thus it is in their financial best interest to use their voting power in a way that benefits the platform, i.e. identifying valuable content and new, aspiring authors. Many of the witnesses have active ongoing projects involving curation and sponsoring community projects. Some of the whales have hired individuals specifically to identify valuable, underappreciated content.

Although the platform is not perfect, the incentive structure has already been through basically 19 iterations, and designing a perfect incentive structure for a complex and large system like this is an extremely complicated problem in game theory.

I hope if you will stick around and post about things that interest you that you will eventually find success.

Best,
Trogdor