If the tests could be done on someone who is asymptomatic, then it could be hugely beneficial.

For anyone already feeling sick... stay home and self-quarantine.

If we could _just_ test everybody regardless of symptoms then we could quickly isolate people spreading it in ignorance. That is the main benefit I could reason out. Probably early detection and better planning as someone with it progresses, too.

The best case scenario for controlling this epidemic is containment through contract tracing. That means we find people who have it, figure out who they got and from and gave it to, and get those people to quarantine. This strategy is failing in the US due to insufficient testing.

You'd hope that someone with a positive test is more careful about isolating themselves and perhaps tells their contacts to do the same.

Besides, if you test someone and he tests negative, you should ideally keep him isolated for a few days and do the test again (cause he might have just caught the virus and be still below detection threshold). It seems impractical.

Depends on their accuracy. A positive illicit test will result in somewhat reduced transmission. A false negative in possibly significantly increased transmission.

Also depends on whether you're testing suspected cases or merely hypochondriacs.

It might reduce transmission if the patient chooses to self-isolate and otherwise would not have.