Because when you make money, someone is paying you for a product or service. When someone chooses to pay you for your product or service, they are expressing a preference for that product or service over the money, and vice versa (you express a preference for the money over the product/service). Which means, that when money is exchanged for a product or service, both parties benefit. Value is created. The utility points of the universe increase.
What you're leaving out, is that "voluntary" transactions have effects on third parties who do not consent, and also true consent of those directly involved requires that they accurately and equally understand the consequences of their actions, which is often not true in real life.
I feel like there is a third issue that I have difficulty articulating, relating to the fact that people are bound to round small probabilities to zero, which means that there is an economy of scale to exploitative behavior.
Sure, those are called externalities. I was simply explaining where the idea that trade generates value comes from. So, given that, generally we presume that trade is beneficial, unless/until someone points out an externality. Which, I certainly wouldn't make the case that there aren't any here, there are. But the way you argue against this activity is by pointing to specific externalities.
> I feel like there is a third issue that I have difficulty articulating, relating to the fact that people are bound to round small probabilities to zero, which means that there is an economy of scale to exploitative behavior.
I don't think what you're getting at has so much to do with probability rounding as it does with 'coordination problems' (http://www.oxfordreference.com/view/10.1093/oi/authority.201...)
If you can come up with a logically consistent definition under which the creation of public property does not itself cause harm - I will be surprised.