From the article, quoting from Nicolas Vabret, an assistant professor of medicine at the Mount Sinai Icahn School of Medicine, who was not involved in the study:
"The mechanisms by which vaccines induce immunity are not necessarily the same as the ones resulting from natural infection," Vabret said. "So the immune protection resulting from a vaccine could last longer or shorter than the one resulting from natural infection."
end quote,
This is significant in that natural infections by the virus provide more antigenic sites in which antibodies can be directed. These sites are not provided by the nucleic acid vaccines, which present only the spike antigens. It will no doubt be instructive to watch what happens with the Chinese vaccine CoronaVax (from Sinovax), which relies on chemically-inactivated whole virus. This is much different and so should be the immune response. Which is better (or worse) remains to be seen.
Curiously there was no mention in the article about mutations of the virus, an active area of research to monitor its changes and how they might negatively impact the current vaccine, and even antibody test kits. This has also raised concerns about changes of the virus from mink farms, and other possible inter-species infections and transmissions.
It would seem rather obvious that memory of the initial virus without mutations is most important regarding long lasting immunity. The only variable in its return seems mutation, since so many other viruses engage in similar activities to reappear - e.g. colds, and flu.. The rate of mutation of this virus over time will almost certainly play a significant role in immunity conferred by any vaccine or a natural infection. Which is all the more reason to stop it quickly. or rather more quickly than we are. The virus cannot mutate without replication. Allow mutations to continue, and the risk to those previously protected, or soon to be protected by a vaccine, might increase. Many of these viruses behave in a similar fashion. Watch the mutation rate.