The "genius" who saves the company from the problems he created, is not the apex negative engineer. The time he spends fixing his own mistakes is the time he does not spend creating new ones. Sometimes his creativity will be limited to a single project.
There is also the type of "genius" who creates an overcomplicated framework as a foundation for a new project, and then leaves the implementation details to lesser minds trying to understand his work, while he moves to start another new project. He is always too busy working on a new project, doesn't have time to answer questions about the old ones. (The problem is sometimes solved by gradually removing his code. In extreme case, the library remains included, to avoid a political conflict, but the code is actually never called.)
> In extreme case, the library remains included, to avoid a political conflict, but the code is actually never called
Ah yes, if the wrong people become aware that the code is never called, better to claim that it happened as too many people had their hand in the pot, and hey, the code is still working, and the tests are passing, so you didn't realize anything was wrong, rather than ruffle feathers and say you didn't want to touch the "genius" code because it was stuck up its own ass.
There is also the type of "genius" who creates an overcomplicated framework as a foundation for a new project, and then leaves the implementation details to lesser minds trying to understand his work, while he moves to start another new project. He is always too busy working on a new project, doesn't have time to answer questions about the old ones. (The problem is sometimes solved by gradually removing his code. In extreme case, the library remains included, to avoid a political conflict, but the code is actually never called.)