We still have not isolated the most important and specific neutralizing antibody (s) (we just group those antibodies that appear to neutralize the virus in the lab) nor the threshold at which most are "protected." So if we do serosurveys in highly vaccinated populations, that data can still be misleading.
This is something that we are able to determine fairly readily for most other pathogens--that allows us to create simple dichotomous antibody tests (yes, antibodies present AND sufficient vs no), but the mutability of SARS-COV2 has really complicated that basic research. And while we know both B-cell memory and T-cell activation is probably considerably longer than the longevity of continually produced circulating antibodies, we certainly don't know the duration of the former. That's why the lack of consensus on WHEN to booster. We can say when levels of neutralizing antibodies have dropped, but if B-cell memory is intact it could simply be hours or days before exposure would lead to sufficient antibody production again (in time to stop the virus before serious illness develops). For those immunologists who believe that the latter is sufficient to prevent most hospitalization/death, they might argue against the imminent need for booster doses or at least a longer duration in the current 5 month-8 month argument. But undoubtedly for some, including the immunocompromised, the b-cell memory may not be "strong" or long-lasting. arguing again for a boosted response sooner.
To be honest (and not wanting to advocate exploiting the undeveloped and thus less vaccinated world), but those more "virgin" populations can provide one hell of a lot of data on the natural immune response, its natural course, duration, and longevity. Hopefully, some countries are able to follow these populations, even while working to secure the vaccines that theyso clearly need.