I hadn't heard of BLAKE3 before (I thought BLAKE2b was the latest and greatest), judging by commits from the repositories of that user it's about half a year old now.
Paper's abstract starts with:
> BLAKE3 [is] an evolution of the BLAKE2 cryptographic hash that is both faster and also more consistently fast across different platforms and input sizes. BLAKE3 supports an unbounded degree of parallelism, using a tree structure that scales up to any number of SIMD lanes and CPU cores. On Intel Cascade Lake-SP, peak single-threaded throughput is 4× that of BLAKE2b, 8× that of SHA-512, and 12× that of SHA-256, and it can scale further using multiple threads. BLAKE3 is also efficient on smaller architectures [like] 32-bit ARM1176
In the introduction, the reason for BLAKE3 is described:
> A drawback of BLAKE2 has been its large number of incompatible variants. The performance tradeoffs between different variants are subtle, and library support is uneven. BLAKE2b is the most widely supported, but it is not the fastest on most platforms. BLAKE2bp and BLAKE2sp are more than twice as fast on modern x86 processors, but they are sparsely supported and rarely adopted. BLAKE3 eliminates this drawback. It is a single algorithm with no variants
Paper: https://github.com/BLAKE3-team/BLAKE3-specs/ (By the way, the download doesn't work for me in the latest Firefox on Linux and GitHub's rendering is plain awful. Not sure if anyone else can reproduce that or if my Firefox is broken.)