[dnbgfher]

I don't believe this was an attempt to find bias or noise from the physical sensors, but rather was looking at the method itself.

The method has some known biases that they were ensuring wouldn't explain the given pattern in the data. Depending on the depth of the simulation, it may have also been able to account for some unknown biases in the method.

Now even under your interpretation, I'm curious about which known object you'd suggest they use.