Great point (love the fallacy of multiple endpoints). But there also may be some "algorithmic" smarts here in terms of Dixon's interpolation/extrapolation of w

Great point (love the fallacy of multiple endpoints). But there also may be some "algorithmic" smarts here in terms of Dixon's interpolation/extrapolation of what cold planetary bodies look like, even with what was known in 1979. Think of an image-generator bot that took some parameters of what was known (distance from Sun, whether it was a gas planet, atmospheric conditions, position relative to the asteroid belt, how size affects features) and kicked out plausible planets. This might be close to what you'd come up with if you made the mistake of forecasting more meteoric impact than actually occurred.