But I think any macro scholar would tell you quite honestly where they think the limits of their models are. It’s like any other scientific endeavor: “this paper on Topic X is our best understanding right now but it’s subject to revisions.”
Similar example from outside macro: there are not many finance professors who will make large bets on the prices of specific assets on the basis of their models, including models in which they have high confidence as being on the current research frontier. That certainly doesn’t mean they are idiots or that they aren’t trying or whatever. It just means it’s really hard.
Imagine a physical science, but make it harder bc you generally cannot do controlled macro experiments (sometimes there are natural experiments if you are very lucky) you can certainly never repeat them AND when your experimental subjects discover that you are trying to predict or understand their behavior they might change it on purpose bc they don’t want to be understood or predicted!
In several ways it’s a lot harder than physics! Murray gell-man said something to that effect once. Something like “imagine how hard physics would be if electrons could think.”