User blog comment:Nedherman1/Arrow Destruction 2.0/@comment-38080588-20190219223520

I'd estimate an approximate FGH growth rate upper bound for \(f(n,n,n)\) to be around \(f_{w^2+w2}(n)\). If made it possible for \(n\) entries, like \(f(n,n,n,...\) with \(n\) \(n\)'s, the upper bound should be about \(f_{w^22}(n)\).