The iterative process of taking risks and realizing profit and loss may indeed be capable of distilling truth from the mass of dissenting opinions in the context of governance - the real question, as I see it, is one of coupling and feedback loops.

If individuals in a system are tightly coupled with the system (participation is high), and if feedback loops are tight (actions produce results quickly), such iteration would be rapid, so the system would evolve quickly.

On the surface, this seems ideal. But many governance decisions require long term planning and have long term impacts which can not be known until significant time has passed. So this rapid evolution would have to occur alongside an increasing volume of such unknowns, and it is not clear to me that such a system would be stable enough to produce better outcomes in the long run than the (extremely bad) systems in place now.

In order for these better outcomes to be produced, from my perspective, there is a massive psychosocial shift (i.e. people would have to become wiser as individuals and in groups) that would have to occur in tandem with the adoption of such a system.

On the other hand, if coupling were low and feedback loops were not tight, such a system would iterate too slowly for its evolution to move us towards a better world in a meaningful time frame, and would get just as tangled up with itself as our current systems tend to.