3-4 week spacing was optimal for getting through the trials and starting vaccinations. I wasn't the best guess for max effectiveness.
It's looking like 16 weeks is much better. The UK took that approach for a different reason - to get as many people with 1-dose coverage as quickly as possible, because they were facing a surge of Alpha, and 1-dose protection was good in that context.
I had dose one on Jan 21, and delayed number 2 until April for the same ethical reason, but it turns out that may be nearly optimal.
With Omicron, it's not clear whether the risk of having only 1 dose for four months is worse than the benefit of better long-term protection.
We've been watching South Africa's Omicron cases increasing, and now we're seeing comparable increases in the UK and EU.
3-day doubling is similar to the near-vertical rise New York experienced with their first outbreak. Keep in mind these are highly vaccinated countries, where something like 80-90% of people are vaxxed or previously infected.
Using deaths from May 1 onward (which is aggressive because that would be infections from early April when vaccines weren't widely deployed).
Granted, this doesn't prove that vaccines are the only factor here, Trump support likely correlates with mask refusal and risky behaviors as well. But what a profound correlation, heavy Trump counties having 2-3x the death rate of heavy Biden counties. Doubly when you remember that Biden took most of the high-density areas.