> While not a perfect measure, proven cases can be used to estimate a persons immunity duration as well as the number of people who had covid but were never diagnosed with covid.

I made a site that does exactly that: http://covid.mremington.co

To estimate true recovered infections from cases, I am using the heuristic defined here: https://covid19-projections.com/estimating-true-infections-r...

Based on T-cell studies, I assume natural immunity to be long lasting, though I may introduce another variable for immunity length.