I agree with all this, although I would phrase the Ethereum part differently. It's all unproven, but thus far, everything that they have said would theoretically work, and tried to implement, turned out to work in practice. Not disagreeing with your point, just the connotation around the word "unproven". The big test for me around ethereum comes when someone pushes the first REAL dApp on to mainnet. If that works out... then things get really interesting.