I think in hindsight we can broaden this argument: human beings do not have the capacity to understand most dynamic complex systems sufficiently to engineer any

I think in hindsight we can broaden this argument: human beings do not have the capacity to understand most dynamic complex systems sufficiently to engineer any particular macro-level outcome, and should instead focus on optimizing specific micro-level components to produce the desired micro-level results. Considering the increasingly common failures of risk-aggregation systems in a number of contexts (not just economic: look at the cascade of failure that triggered the 2003 East Coast Blackout), we should be increasingly conscious of this as technological improvements allow us to act unilaterally at increasingly large scales.