WireGuard and OpenVPN are good protocols. They're battle-tested, widely deployed, well-understood. For a network where "any encrypted tunnel" is enough, both work fine. So why a different protocol at all?
The first difference: appearance to the network. WireGuard and OpenVPN have well-known signatures. A network that looks for "is this a VPN" can find both within a handful of packets. FROST/1 is shaped to not present those signatures. We don't publish how — the how is part of the protection. What's user-visible is the property: on networks that block obvious VPNs, FROST/1 has a substantially better chance of staying up.
The second difference: directory and routing. FROST/1 is built around an open node directory and per-node selection by the user. The user picks the exit; nothing rotates without their input. This is closer in spirit to OpenVPN's classic config-per-server style than to WireGuard's typical commercial deployment, where a "magic" server is chosen for you.
The third difference: identity. FROST/1 uses voucher-based access at the application layer. The protocol itself doesn't care; the access model around it does. WireGuard and OpenVPN use cryptographic keys per peer, which is fine in self-hosted deployments and creates a long paper trail in commercial ones (key issued to whom, on what date, paid by whom). The voucher model collapses that paper trail to a bearer code.
What FROST/1 doesn't do: it doesn't claim to be faster, it doesn't claim better cryptography in any meaningful sense (good crypto is not scarce), it doesn't claim to be simpler. WireGuard is famously simple, and FROST/1 isn't trying to compete on that axis.
If "any tunnel" works for your network and you want maximum simplicity, run WireGuard. If you need a tunnel that doesn't look like a tunnel and don't want to operate the infrastructure yourself, FROST/1 is the differentiator. The choice is shaped by the network, not by abstract protocol preferences.