At least four years ago, Dan Bernstein argued residue number systems provided parallel speed ups across all VDF designs: https://nitter.net/hashbreaker/status/1131192651338829825
An adversarial advantage of 2 never sounded realistic of course, never mind all those silly academic papers with adversarial advantage of 1.
I suppose this shoots down protocols using an adversarial advantage of 10 or even 100 too?