Understanding performance and result trade-offs of various sorting algorithms is sometimes useful. But usually not.
Understanding how the nagle algorithm works in TCP is sometimes useful. But usually not.
Understanding how the A* algorithm works is sometimes useful. But usually not.
Understanding how zlib works is sometimes useful. But usually not.
That is the beauty of our field. We can build larger and more complex systems because others have laid the foundation. Actively seek out and learn relevant existing algorithms and libraries when you need them.
All of this stuff may or may not be useful to know. However, it is largely irrelevant when it comes to getting shit done in most bits of software. Sometimes, ignorance is blissfully productive.
What makes you happy? Screw the rest.