The article uses the testing of asymptomatic people who've potentially been exposed as a jumping-off point to push the authors' proposal for testing asymptomatic people regardless of known exposure. These are of course completely unrelated in terms of actual benefits, practicality, etc but it treats one as though it were an extension of the other. It can even make sense to test selected high-risk or high-danger people without known exposure whilst not testing people known to be exposed in general; I know the UK is doing this to a certain extent for example.

As for testing of potentially exposed people without symptoms, it has its issues as the CDC pointed out. The test just can't reliably detect people who're infected but still in the incubation period of the disease, so testing them will tell a bunch of them that they're negative and some of them will think they're in the clear and go about their normal business. I think New Zealand had a incident recently where a health worker in a city they were hoping to stop the disease spreading to tested negative, went back to work, and then tested positive again later after developing symptoms and working for a while whilst infectious, though that one was probably hard to avoid. It's also not terribly commonplace, especially as a companion to mass testing everyone with symptoms rather than an alternative - I know the UK in particular, which is the other country with a big mass testing program and probably the only one with relatively working contact tracing too, doesn't do this as a matter of policy. It also consumes a bunch of tests that would be more usefully used dealing with the test backlog - finding potentially exposed people should work a lot better if you find out they're potentially exposed sooner, and the main tactic is getting them to self isolate.