I agree that this can be done with side chains, and this is part of the reason for this design. It adapts to side chains or non side chained altchains. I am of the mind thinking there will be many blockchain based networks some day each with a specific niche. If Bitcoin was the standard P2P payment network globally, I would not be surprised. But I would be surprised if developers decided to only build new blockchain models off of sidechains.
There is a steady stream of work saying POS variations can be strong enough to run blockchains and have a tough time believing that Bitshares, Ethereum, NXT, and Peershares et. al. are just going to get made irrelevant by all the POW side chained alternatives. So I am planning both for side chained and non side chained blockchains in the future that have their own economic systems.
Furthermore, Lighting Payment Channel based networks could provide a near real time off chain layer for smart contracts based on the basics outlined above alleviating blockchain bloat on any one blockchain that is not subject to the txnID malleability problems.
These are all contingencies I am balancing for.