You’ll understand the bug better. This means you can write a better bug report, which will help Apple fix it more quickly (meaning you may get the fix more quickly).
They’ll understand the bug better. This, too, helps Apple fix it more quickly.
You may find that it is not a bug in the API at all, but that you were misusing it. Perhaps you were using something on a thread that you shouldn’t have been, or expecting some argument to be used a certain way when it’s actually used differently.
In this case, you may be able to use the API after all, saving you the time you would have spent hacking around a non-bug. This also saves them the time they would have spent triaging and eventually responding to a non-bug.
If your misunderstanding was borne out of poor documentation (misleading, inaccurate, vague, incomplete), you can file a bug report about that instead. Then the documentation gets better and future users of the same API avoid making the same error you did.
A test app isn’t appropriate for every kind of Radar, but when it is, including it helps everyone.