It's an educational project. No need to put it on blast over that. CE/EE students can buy a board for a couple hundred bucks and play around with this to learn.
A hypothetical ASIC implementation would beat a CPU rather soundly on a per watt and per dollar basis, which is why we have hardware acceleration for other protocols on high end network adaptors.
Personally, if I could buy a Wireguard appliance that was decent for the cost, I'd be interested in that. I ran a FreeBSD server in my closet to do similar things back in the day and don't feel the need to futz around with that again.
I agree that if the goal is to be educational, it's an excellent interesting project. But there is no need to make dishonest claims on their web page like "the software performance is far below the speed of wire"