The SHAKE family of functions (and cryptographic sponges in general) are very cool, but like the article says, SHA-3 is weakened by failing to focus on performa

The SHAKE family of functions (and cryptographic sponges in general) are very cool, but like the article says, SHA-3 is weakened by failing to focus on performance. If SHA-2 is sufficiently secure (which looks more and more the case) then a faster SHA-2 would have been a better way of looking at what was wanted out of SHA-3. Faster primitives means more PBKDF rounds, more usage of hashes to secure underlying data.