Yeah... -- reading the wireguard paper[0] now, and my cursory analysis ("oh this is like that time that I...") of how it works/what it needed to accomplish is suuuuper naive.
The mathematical foundations and power/elegance (and often even simplicity) of cryptographic primitives make it really easy to get the completely wrong idea about the difficulty and disturbing fragility of actual cryptographic engineering. Happens to pretty much everyone.
[0]: https://www.wireguard.com/papers/wireguard.pdf