Because the data shows they don't matter.
This is the same thing as deaths due to covid vs. deaths while positive to covid. Sure some people might have been miscounted as covid deaths but actually died due to cancer/car accident/whatever. Despite this the excess deaths over the past two years is much higher than the number of covid deaths; thus showing that miscounts must be a small minority which, in any case, is absolutely dwarfed by excess deaths not counted as covid deaths. It might even grow a little as testing improves but it still remains mostly irrelevant.
Likewise it may be that some people do not need a vaccination. However, we have a 30/70 split in number of covid hospitalizations, with 30 being vaccinated people, in countries where the split in the normal population is 90/10. This means that despite some unvaccinated people having had prior infection the vaccine reduces hospitalizations by 20x. Given this data the most effective strategy is to just vaccinate everyone without what is effectively a pointless distinction.
As the number of unvaccinated but protected people will grow (through a combination of more infections, more vaccinations and more deaths), the proportion above will revert to 90/10 and unvaccinated people will not be an issue anymore. For now however analyzing the proportion of naturally immune people among the unvaccinated is, again, mostly irrelevant.