The current network design does not scale at all, we need to either change it or build a micropayment system around it 
No.
See previous posts in same thread, bitcoin banks will handle that much better than any blockchain.
Where can I find a careful performance analysis of bitcoin? How does it scale with millions or billions of users, in terms of such things as transaction speed, computational and memory requirements per node, network bandwidth requirements, ...?
An ideal system would allow a billion human users all to issue 0.0001 BTC transactions at the same time, and have all such transactions (for those users whose local network and computer resources function adequately) to complete in under a second. I presume that neither bitcoin nor any other system we are aware of can reach this ideal. But I doubt bitcoin can succeed unless it has a performance analysis, preferably of peer-reviewed research paper quality, and a clear strategy for dealing with whatever manner and way it falls short of the impossible ideal I described.
But I'm new here; so likely such a performance analysis exists and I just have not noticed it yet.