It's a different model from reward pools for sure. It puts a high priority on witnesses to set APR appropriately (so vote accordingly!). And the haircut of 10% may be too low or have been a bad idea in the first place (if you know the history, it was made up by Steemit on basically no analysis). So overall I agree with you on the high potential for significant speed bumps. However, I am not sure it will fail.