I don't think that is an appropriate generalisation, because you are drawing a similarity between the game theoretical outcome of properties on systems of ratio

I don't think that is an appropriate generalisation, because you are drawing a similarity between the game theoretical outcome of properties on systems of rational actors and the outcome of properties due to random physical occurrences ('cosmic rays' etc...) and unintentional irrational behaviour ('human error'). While both can result in outcomes that are collectively undesirable, I don't think useful inferences can be made by making a vague generalisation by pointing to two different classes of system (especially not without defining the generalisation unambiguously and proving rigorously that it is always true).