Since it's possible for witnesses to generate blocks without solving a proof of work, what is the reason for including one miner generated proof of work block per round instead of just relying entirely on the witnesses to generate/secure the blockchain?
Is the chain somehow more secure due to the inclusion of the miner generated block? What's the advantage to including a miner implementation at all within a DPoS system? Thanks!
I believe I've found at least one answer to my own question....
From the Steem Whitepaper, it looks like solely relying on elected witnesses to produce blocks would open the blockchain up to censorship. The witnesses could collaborate to censor transactions from a certain user. Therefore, the one randomly selected witness from the pool of vote getters and the proof of work selected witness, allow anyone to eventually get a turn a block production. And when their turn comes around, they can always include their own transactions, thereby reducing the effects of the censorship.
The proof of work elected witness allows someone who doesn't have enough votes to jump the line via using computation power, and the randomly elected witness amongst vote getters can jump the line by purchasing Steem Power and voting themselves up.