You could spend all day real time looking at the fact that few of the mortgages in your CDO are failing, and then when they fail systemically, you could watch in real time that a lot of assets that were thought to be pretty robust are not.
Your argument seems to be simply that radically more data would enable better predictors. Whether that is true or not, it's unclear whether bitcoin et.al. could help much with delivering that data. If we agree to a contract that I will pay you 5 bitcoin tomorrow in return for 4.9 bitcoin today, that contract is not publically available on bitcoin. Nor is, the fact that a certain money stream is part of a mortgage.
So unless you force every contract to be public, blockchain technology will not help you that much. And if you force every contract to be public, then you don't need a blockchain.