Miners almost never finish their mines for the estimated cost and in the estimated time period. Also, once the mines are finished, they don't tend to produce the amount of resource that was estimated.
This inability to accurately project means that miners get in trouble with debt financing. (Which is why streams are attractive.)
Taleb would recognize the problem with the mine completion estimates: convexity. The model error is always in one direction: behind schedule and over budget.
One alarming thing about the talk was Watson saying that stream financers had historically tried to squeeze as much value from the mines as possible. Sounds great. Now he is trying to create more "win win" deals. I'm wondering whether he is just pitching to a room full of miners, or whether the economic recovery has made it harder to find distressed miners that he can get a huge vig from?