Beyond the doomsday economics of “proof-of-work” in cryptocurrencies [pdf]

I wonder why the paper didn't consider a constant block reward (i.e. pure linear emission) as a way to mitigate the issues caused by relying on transaction fees for long term security.