This is indeed two sides of the same coin. A good reminder that you need to pick the right approach depending on the context.

Interesting point of view. I'm not sure I fully agree with the classification but it gives something to mull over. For sure the less reliable your estimates the more padding is needed to have some predictability.

Very interesting discussion about decision making and forecasting. I'm discovering Tetlock as he is mentioned and his findings are interesting in their own right as well.