From the data it’s not clear that shelter in place orders have had anything other than a marginal effect.
I do some contracting work for a British company, and had some extremely unrealistic deadlines looming in late February.
For this reason, I was keenly watching their government's announcements at the time and was amazed as weeks went by and my country's restrictions got tighter and tighter while the UK didn't budge.
Now, in May, it's not exactly apples to apples but they're now the hardest hit (deaths-wise) in Europe, and my country has dropped down below 1000 active cases (and decreasing some 50/day).
You could probably substitute my country for almost any other early mover and the UK for almost any other late mover and arrive at the same result. Even across different pandemics.
"Moving", here, means locking down.
Given that, my personal biases lead me to think that lockdowns probably work. I have to acknowledge that there's a moderately high bar set for data which disproves this thesis, because it's just so damned intuitive.
This doesn't meet that bar.
In reality though, I feel like in comparatively sparsely populated places like Sweden, New Zealand, etc. there's not a lot of difference at this point between a supermarket with significant social distancing imposed and a cafe with significant social distancing restrictions imposed.
It's once you get to the densities of London, NY, Madrid, Rio and Paris that the difference starts to become important.
From the perspective of reducing the spread, this should mean that many of the US states should have been able to emulate Sweden's approach (inform citizens and trust them to cooperate - up to the reader to decide if this translates between Scandinavia and the US), and many are trying to. I guess we'll see, but I think it's pretty clear that if you're New York you don't have that option on the table.
But it's also about the relative danger. Hospitality staff in South Australia right now, were that state to return to "business as usual", probably aren't being asked to risk their lives. That is and has been a different story in different states and cities around the world.
It was fairly evident that every country will end up in a lockdown except that the countries where it was applied early did not encounter huge loss of life.
What data are you referring to that suggest shelter in place orders could be having only a marginal effect?
We're living in a hyper-partizan country right now and everything published about this virus is going to be taken by both parties as a way to support their case and attack the other. Until there's another civil war or the country decides to break into smaller nations this is only going to get worse.