> But if you needed some kind of custom sort function to serve features within a greater application, you are already going to know that your sort function works correctly by virtue of the greater application working correctly. Testing the sort function in isolation is ultimately pointless.

It is entirely possible for a sort function to be just one component of the functionality of the larger code base. Sort in specific is something I've written unit tests for.

> As before, there may be some benefit in writing code to run that sort function in isolation during development to help pinpoint what edge cases need to be considered, but there isn't any real value in keeping that around after development is done.

Those edge cases (and normal cases) continue to exist after the code is written. And if you find a new edge case later and need to change the code, then having the previous unit tests in place gives a certain amount of confidence that your changes (for the new case) aren't breaking anything. Generally, the only time I _remove_ unit tests is if I'm changing to a new implementation; when the method being tested no longer exists.