So we're stuck looking at real world data sets.
I don't have any issue with analyzing the data. As I said, show me the multivariate analysis. This article hasn't even done that.
If it turns out that the article's hypothesis is correct but only because it forces people to go outside and get sunlight and the sunlight is healthy and... we would have no way of knowing. Because one policy by one country only provides one data point.
I keep saying everything you're saying is true, and yet this does not change the fact that the article is engaging in the broadest level of speculation possible with no falsifiability whatsoever. The article may as well claim "naming your country Denmark causes a decrease in COVID hospitalizations" because, after all, that hypothesis is consistent with the data!
You said
This is about what an entire *country* is doing
Denmark does ten thousand things that no other country on earth does. As far as I'm concerned, that makes it a single data point. Which of these unique-to-Denmark policies, and to what magnitude, are they responsible for their different outcome? All we can do is speculate. No magnitude was claimed or is even known. It's a hypothesis. That's it.
For example, there are no controlled experiments about how much lead in the blood effects children for obvious ethical issues
Except there are. We can do multivariate analysis and isolate random environmental factors that cause a family to move into a house with a higher or lower level of lead paint, measure how much lead is in each resident's blood, and that suffices as an independent variable.
The article hasn't done anything like this. In fact, if I did the same quality of analysis as the article assumes, the result would be that masks cause COVID deaths. Why? Because cities typically implement mask mandates as infections go up, and deaths typically lag infections by a couple weeks. Can you see why I'm skeptical now?
Even more scientifically, how this usually works is we start with a hypothesis that pool fences (or refinishing a house with lead paint, or whatever) saves lives. Someone gets a grant to install as many pool fences as they can, and then they follow similar families who didn't get the grant because you can't help every person in the country at the same time and that's just life. And then you find out that families with pool fences have fewer drownings.
My city just did an experiment like this with a newer kind of walk signal at a few random intersections in the city, and they saw a 4x reduction in pedestrian collisions. It was so successful they're trying it at many more intersections.