Ping.
I am highly skeptical of antibody tests. This is because, in my practical experience, I almost never have seen an antibody that recognizes a single antigen. This is because antibodies recognize antigens on the basis of physical characteristics that may be shared among many proteins.
I read a few days ago of how serology (antibody) studies were used to study a particular virus for decades, until techniques to study the molecular composition of the virus were developed. It turned out that what was believed to be one virus was actually three related viruses.
Similarly, Covid-19 is one of a group of coronaviruses. Some researchers have already noted cross-reactivity between Covid-19 (which they call SARS2) and SARS. [https://www.biorxiv.org/content/10.1101/2020.03.15.993097v1] Covid-19 is already known to infect cells through the same pathway as other coronaviruses; this fact alone suggests that antibodies against other coronaviruses will also react against Covid-19, making serology testing useless as an indicator of past infection.
What this means in terms of the disease itself is still unclear. Can cross-reactive antibodies protect against Covid-19 disease? We have seen that Covid-19 illness in young children is not as serious as in older people. Young children who are notoriously prone to catching frequent colds may have already had recent infection with one or more of the circulating coronaviruses. On the other hand, antibodies sometimes enhance the virulence of a pathogen; this is seen in the case of FIP, a fatal coronavirus infection that affects cats.
The CDC warns that positive antibody tests in an area where Covid-19 activity is low are likely to be false positives. And many of the claims that Covid-19 is widespread (despite an absence of confirmed cases) are based on serology testing.
While serology testing has value in Covid-19 research, I would not use it as a means of determining past infection. The only accurate ways to determine infection is through virus characterization using RT-PCR or virus culture. I do not know how long virus particles remain in the body after the patient recovers from the infection, but it may be weeks: news media reports indicate that Covid-19 virus was found in Floyd George’s body at autopsy, and he was known to be infected in (IIRC) March. So it might be possible to recover virus particles after the fact to determine past infection.