> It seems that the cost of preventing counterfeiting, and other comparable attacks against the conventional financial system, is much lower than than the cost of the proof-of-work calculations done in bitcoin.

What do you base this assertion on? I have no idea what the US spends annually on preventing counterfeiting. Do you?