The points the author makes are great. Nevertheless : point (1) is anti-scientific, and (2) is pessimistic.

(1) is anti-scientific: Suppose I said “we will never predict weather! It is too complicated. Imagine what is required. Each cloud has a thousand different factors acting upon it. To get a clear picture of weather, we need to predict what happens to every cloud at once! Impossible!”

The way out, of course, is: yeah it’s hard but you can approximate. We don’t have to get our happiness measure perfectly right (we never will), but it should be usable.

But yes, if the metric is fuzzy we shouldn’t design hyper-detailed policies based on what it says.

The scientific design of policy requires an appreciation of the uncertainty of the measured values. More uncertainty means vaguer implications for policy.

Point (2) is pessimistic. The current culture may not support scientific implementation of policy but that doesn’t mean it is wrong to try to get there.

Finally, all these terms are relative. The two authors are probably just saying that “the more accurate our metrics for human well-being become, the better we will be at making policies that support human well-being.”

This seems uncontroversial. Note that GDP per capita is just such an imperfect metric. It has been very helpful indeed. It is believable that we will someday design new metrics of well-being that can be even more helpful than GDP.