Based on a few things I've gathered from news and the TWIV podcast series, this is what I think might happen:
The virologists on TWIV think that, based on the four other known human coronaviruses, SARS-COV2 will eventually take on the characteristics of the others. The other coronaviruses are one of the viral causes of the common cold-mild illnesses with extremely rare secondary infections.
There are a few documented cases of people who have been reinfected with the coronavirus. The infections have been with a slightly mutated isolate (only picked up genetically). One of the several documented cases had a more severe infection the second time. In the other cases, the disease was much milder. Immunity is heterogeneous. Those studying the disease currently don't know what elements of the immune system are most at play, as far as which antibodies are the most effective, what is the roles of T cell immunity, dendritic cells, etc.
TWIV.TV episode #664 presented an elegant paper looking at variations of antibodies to the known coronaviruses over time. They used a cohort of HIV infected men who had their blood sampled every three months since 1989 (though there was a 4 year gap early on). They went back to the many years of frozen sera from ten of these patients and measured antibodies to these common coronaviruses. They were able to document that detectable antibody lasted about 6 months after an infection. Immunity starts to wane to these viruses after 6 months, so one can get it again. But when you get it again, the illness is milder, sometimes asymptomatic, even in these HIV-infected patients.
The hope is that a vaccine (two doses) will give the vast majority of us significant immunity, and that second, third, and subsequent exposures will not cause significant clinical disease. The vaccines may not prevent infection, but they may prevent disease from the virus. That is the ultimate goal anyway. If each infection boosts immunity yet doesn't make you sick, so much the better.
It is interesting that infection in children is in general much milder. So, in the future, we may be vaccinating people against this at a certain young age (once safety studies are done), and then with future exposures, the vaccinated don't get significantly ill.
During the 1918 flu pandemic, older people got severe illness much less frequently than the younger people. The peak age of those who died was 28. There was significant illness in children and in young adults. Why? Probably because the older folks had survived a previous flu pandemic and already had partial immunity. There was a flu pandemic that went worldwide in 1889-1892 and killed about one million people.
This is going to take awhile. But we will get there. And it will probably end before we notice that it ended.