The problem with your solution is that implementing it, while improving the quality of the platform as a whole, would not be in the best interests of the developers. Simply put, people are only using Steemit to make money. Nobody is saying "Wow, look at the quality of content on Steemit!" The ONLY draw to the platform is the profit possibilities. By fragmenting voting into niches, there will no longer be posts earning tens of thousands of dollars.
My proposal would perhaps turn 1000 Steem posts into perhaps 100 Steem posts, but likely Steem would thus be 10X higher in price because of the greater participation and utility. So nothing lost and only gained.
You're falling under the inaccurate assumption that a 10x better system will = 10x the investments into the system. The people making big bucks off of Steem don't want to fix the system. It's broken in their favor. While new investors may see such a change as a better chance of earning, the people holding and distributing Steem will see it as absolute loss of profit. If whales are dumping, even if the fish are pumping, the price is not going to hold, and certainly won't appreciate 10x.