What does the decline of serum antibodies after infection mean for B cell immunity, exactly?

About a month ago there was a small media blip about a report in the New England Journal of Medicine that neutralizing antibodies to SARS-CoV-2 decline significantly within a matter of months. Published op-eds ranged from "a vaccine won’t work, why God WHY" to "actually this isn’t that weird, stop freaking out." Some key details from the articles I read at the time:

1. We still got T-cell immunity, and there are some signs this could last longer for COVID-19
2. Antibodies decline as infections clear for other illnesses.
3. Remaining memory B cells are (usually) primed to churn out more antibodies upon reinfection.
4. Nevertheless, the decline of antibodies is less protective and raises concerns of decay of immunity (an opinion in the original report)

I want to ask about the B cell side of this in order to better understand the somewhat conflicting messages and opinions.

1. If memory B cells worked as advertised, how would serum antibody levels even be a measure of immunity? It would seem that antibodies dropping to negligible levels wouldn’t matter if your memory B cells can make them on demand.
2. Is there some benefit to having antibodies in your blood independent of your memory B cell function? Is it somehow quicker or vital to prevent reinfection or recurrence of serious illness?