I'm really happy that GitHub supports rendering ipynb files so nicely. Makes it easy to glance at repos like this without cloning and firing up a Jupyter notebook.
It seems like the algorithms aren't that complex, after all they have to be completed in one day. There might be more value in something like, 12 algorithms/side projects a year. Enough time is had in a month to actually develop something meaningful, not necessarily an entire side project but a deep understanding of a specific algorithm.
What's the benefit you get out of implementing an algorithm that you don't get out of reading carefully its correctness and complexity proofs?