> 2. Their paper doesn’t claim anything is wrong with 25519. They’re just proposing a random Edwards curve alternative to 25519 Which brings me back to the million-dollar question: why do I want this?