The skinny on how antibodies and antibody testing work

I thought this article in Smithsonian Magazine gave a good overview of how antibodies and immunity work, how complicated it all is, and how it might play out in the case of SARS-CoV-2, the virus responsible for the COVID-19 infection. It does not make good reading for those looking for a quick and final end to the pandemic we currently find ourselves in.
When a person gets infected by a virus, the immune system first mounts a shortlived "innate immune response" against anything that looks like a foreign body. Over time (a week or a few weeks), a more specific "adaptive immune response" is developed, which can recognize and respond to the unique features of the invading microbe. In this second wave, immune cells called B cells manufacture antibodies against the virus, which is what the antibody (or serology) tests we have all read about are looking for.
However, some of these antibodies are full-blown "neutralizing antibodies", which curb a microbes's ability to latch onto and enter cells, while others simply flag the existence of foreign germs so that other parts of the immune system can hopefully go to work on them (neutralizing antibodies are obviously the ones we want to see). Some of these helper antibodies, for reasons that no-one seems able to explain, can even operate in an antagonistic way, chauffeuring active viruses into healthy cells, which can serve to accelerate the infection ("antibody-dependent enhancement"). Antibodies typically only have a relatively short life-span (a few weeks or months), after which they are cleared out of the body on the grounds that they are probably no longer needed, although the body does retain some B cells so that if the virus is encountered again it can start to generate neutralizing antibodies more quickly.
Furthermore, we need more than B cells for an effective antibody response. We also need T cells which, among other functions, help young B cells grow up into effective antibody-making machines, and can also trigger infected cells to self-destruct in their own right, and kind of patrol tissues after an infection to make sure that germs cannot establish a new foothold. Unfortunately, T cells tend to be located in hard-to-reach locations like the lungs, and so, even though they are an essential part of a robust immune response to a new infection, they are very difficult to detect and analyze.
Setting aside the possibility of false-positive and false-negative errors that beset all immunity testing, the serology tests that are being developed as we speak typically just search the blood for antibodies that can detect the spike proteins of the SARS-CoV-2 virus (what it uses to attach to and infect human cells). They don't detect whether the antibodies are neutralizing or just the less effective helper antibodies (that requires additional analysis). And bear in mind that some people that do have neutralizing antibodies have been shown to still succumb to the COVID-19 infection (presumably because other parts of the immune system are also needed).
So, the bottom line is that antibodies may not offer complete future protection from the virus (a "sterilizing immunity" response), and antibody tests are far from definitive or foolproof. Without full protection, reinfected patients may experience milder symptoms, or no symptoms at all, but still be able to transmit the infection. Some, however, may experience similar, or even more intense, symptoms the second time around. And how long any immunity will last is another unknown at this point. A new study at the University of Amsterdam suggests that immunity to coronaviruses in general may only last six months, making the idea of "immunity passports" of little value, although it is not yet known how COVID-19 may fit into this framework.
It's certainly interesting and instructive that doctors and virologists who have had COVID-19 are still taking all the same precautions they were taking before infection...