Yes that's right, people often miss the trees for the wood too.
Especially when dealing with complex systems with many parts.
For example, consider technological progress or the stock market. There are so many interacting parts that people seem to not think about the parts. They merely try to predict the emergent outcome as if it were an entity by itself. And they run into estimates that are far off.
My favorite example are economic aggregates like GDP, GNP, Income equality, Wealth distribution, Consumption, Production. I see people fitting increasingly complex auto-regressive models, testing for co-integration, correlation with no regard for the underlying reality.
If a whole economy can be manipulated by a single number like interest rate, I wouldn't be living in it. That economy would be too fragile to prevail.