Yeah, this is a common pain point in CRA projects. I’ve had success suppressing logs during passing tests by mocking console.log in a setupTests.js file.
You can override it to only print when process.env.NODE_ENV === 'test' && someGlobalFlag.
Then in your test failure handlers, flip that flag on. It’s a bit of a hack, but it really cleans things up and keeps focus on broken tests.