"Two new studies demonstrate how severity of disease is predictive of longer-lasting antibody production and detail how immunity wanes over time but may exist for up to 7 months."
https://www.cidrap.umn.edu/news-perspective/2020/10/studies-...
https://www.cdc.gov/coronavirus/2019-ncov/hcp/duration-isola...
There's also evidence of re-infection
https://www.thelancet.com/journals/laninf/article/PIIS1473-3...
There is evidence that reinfection does occasionally happen, but none that it happens frequently enough to worry about. The paper you've linked mentions four cases.
> The paper you've linked mentions four cases.
Those are confirmed cases. That doesn't mean there's only 4 total.
> Those are confirmed cases. That doesn't mean there's only 4 total.
I'm sure there are more, but how many? There have been hundreds of millions of infections (according to the WHO). If there are a few dozen or even a thousand reinfections among them, it's not a phenomenon we need to worry about.
We could assume that reinfections are common even without any evidence for or against, but then we are playing something akin to Pascal's wager. Also, I think if reinfections were common, by now we would very likely have evidence for that by now.
And of course there are going to be individual cases of reinfections with 55+ million cases.