Nice post, thank you for keeping it short.
I am sad and concerned about this.
The way the white paper is written, its contents does not bring me any joy...
As a developer: if we use a blockchain it is also to communicate with each other. Now, you have to go to blockchain X,Y,Z because the needed information is only inside, you just have this "signature" at the root blockchain. It will be not very nice to work with.
The Map/Reduce framework works well only for dataflow style applications with few states. Also, someone will have to pay for all those additional blocks.
I see a block chain has a commodity for storage and/or computation, who will win: cheapest transactions/storage/computation, simplicity to build applications, throughput, scalability, security.
I do not see that in Ethereum in the near future. Sure it will keep momentum thanks to ICOs, and many actors pouring money into. But this is not an exciting future for me.
Hmm. Yes I can understand your concerns.
I guess we will have to wait and see how this proposal would turn out in real life, and how the community reacts to it.