I can start and others may add.
Right now what we know is that most people who have been infected will develop some kinds of antibodies. The amount of antibodies and the quality of these antibodies are not known for sure. They may vary from one person to the other. They may vary on the type of exposure a person has had, whether they've had a severe disease or whether they've had a mild disease, in which case they might not mount a very big immune response.
The jury is still out in terms of the quality of the immune response as to whether it is really very protective against reinfection. We can assume that probably it would be somewhat effective against reinfection.
What is less known is the durability of this immune response, of these antibodies. From a recent study that was published last week, immune responses were evaluated over a period of 35 years in 10 individuals, and all their coronavirus infections were mapped. The conclusion was that the lasting immunity against a coronavirus infection was between six to 12 months.
This is a problem because it's not a very long-lasting immunity. If we want to rely on herd immunity, it would take forever for a society to be 70% totally immune against these types of viruses. That's where we are right now.