I suppose the question is, if such a system works most of the time does it matter? By trying to only find solutions that are 100% game proof we may end up with generally subpar systems in reality.

Further if such a system were found I am almost certain some person somewhere would figure out a means to exploit it

Bad human behaviour can never be completely eliminated. All you can do is try to mitigate against it where you can but it isn't always possible.

This is the difference between how many people who work in computing think vs what we see in biological systems.

Biological systems work optimally most of the time and in most situations, whether they work in exceptional circumstances or not is somewhat irrelevant.

I'm not saying we should ignore potential problems or risks only that we should not lose sight of what is important and the vast bulk of common circumstances, merely to mitigate against exceptions.