Exceeding low estimates - a success or still a failure?

From Assessing My COVID Expectations by Bryan Caplan.  Much success depends on effective forecasting on select issues.  Forecasting is cognitively taxing and diversionary from more mundane activities which necessarily must be tended to but a usefully accurate forecast on something germane can yield significant dividends.

Caplan is post-event assessing actual performance during the Covid-19 pandemic against his expectations which requires a little mental limberness.  We are accustomed to thinking in absolute terms.  If you want something to perform at a certain level, that is what you measure.  This machine is supposed to produce 100 widgets per hour is a forecast in absolute terms.  A more Bayesian approach takes into account contextual expectations.  

Yes, it is supposed to produce 100 widgets per hour but it is a fiddly machine and changes in the quality of inputs lead me to believe it will only produce 20 widgets per hour.

If it produces 50 widgets, it far exceeds the low expectations based on a Bayesian assessment and far underperforms based on a more theoretical expectation.

Caplan looks at the performance of Government, Population, Business, the Pharma Industry, and the FDA in particular.  

My paraphrase of his assessment:

Government - He had low expectations and Government underperformed against those expectations.

Population - He had moderate expectations of people's response to the pandemic but they dramatically underperformed against expectations.  His chief surprise was how easily people were stampeded into hysteria and how resistant they were to reality checks of that hysteria.

Business -  He had high expectations and business slightly exceeded those expectations.

Pharma Industry - Given the significant possibility that a vaccine could not be developed at all, he had reasonably low expectations and they far exceeded those expectations, producing multiple workable vaccines within a year which were then manufactured in bulk and widely distributed quickly.  

FDA - Very, very low expectations but their performance far exceeded those low expectations. 

I broadly concur.  I would assess government performance even lower.  Had we but adhered to the existing pandemic plans, we would have avoided prolonged shutdowns and indiscriminate masking debates leading to measurable better outcomes.  By doing nothing other than what was already set in place based on past experience, outcomes would have been significantly better.  

In terms of the population, I am sympathetic to a pass.  When government, academia, and the media are all lined up propagating hysteria, it seems unreasonable to me to expect that the average busy citizen should stand as a rock against the storm of misinformation.  

Business and Pharma - agree.

I am curious why Caplan focuses on the FDA rather than the CDC.  I can sort of see how he could get to his FDA conclusion since he expected the very worst.  I would be curious though, to see his assessment of the CDC given its centrality to the issue of response to pandemics, how they arbitrarily through out the established playbook, went with their guts against the science, flip-flopped, worked agendas not in the interest of the public, propagated known untruths, and hid relevant conflicts of interests. 

I had moderately low expectations but their emphatic departure from science and professional norms seems to warrant a bad performance against moderately low expectations outcome.  

Regardless, I like the Bayesian approach which goes beyond absolute assessments.