The paper confirms previous evidence of the generation of memory T-cells, maintenance of antibodies for several months and also cross-reactivity with other coronaviruses.

The highlight of the paper is the systematic study and analysis of memory cell populations. They found a subpopulation, called memory T stem, which is able to activate against virus peptides and is a long-term reservoir cell, as it is able to activate on sight of the virus and has been described as being able to give rise to a long-lasting memory.

Analysis of these cells suggests that vaccination, regardless of whether there is a drop in antibody or even T-cell production, induces the generation of a reservoir of cells that can maintain memory for long periods of time. How long they can persist is not known specifically, but for other pathogens they can be maintained for decades.

It would indicate that, at least in the general population, the immunity generated by the vaccine is sufficiently effective and, most importantly, the memory it induces will be long-lasting.  Therefore, as far as we know, a third dose would not be necessary in the general population.

In immunosuppressed or immunosenescent people, the third dose is actually to complete their vaccination schedule, as they did not respond well with two doses.