>we could do (very similar) analysis on "normal" bank accounts
I don't think so really, the fact that the ledger is public is really the thing that makes this is a credible model.
In the absolutely general sense of "if all banks around the world opened all their books and banking secrecy laws didn't exist, and if we somehow had the ability to trace through cash transactions" then sure, theoretically (and again not accounting for the absolutely vast difference in scale between the volume of transactions in the real world vs the blockchain), but none of that is even remotely probable.
OK, sure, in about the same sense that world government is possible: hardly anyone is asking for it and there is a tremendous level of investment in the status quo by all the people with political power.
I'd argue that the problem isn't secrecy laws but just general incompetence and unwillingness.
Sure, banking secrecy laws could be a problem for large-scale frauds totaling millions, but smaller-scale operations typically stay within the same country where the law most likely already allows this kind of tracing, but it's so unefficient that by the time it's tracked down the money is already gone for good.
I don't think so really, the fact that the ledger is public is really the thing that makes this is a credible model.
In the absolutely general sense of "if all banks around the world opened all their books and banking secrecy laws didn't exist, and if we somehow had the ability to trace through cash transactions" then sure, theoretically (and again not accounting for the absolutely vast difference in scale between the volume of transactions in the real world vs the blockchain), but none of that is even remotely probable.