The problem comes with relativity. Suppose that an observer determines that event "A" is in some way the "cause" of event "B." If FTL communication is possible, then relativity allows the existence of another observer who sees "B" as occuring in "A's" past, i.e., as the effect having preceded the cause.
I think it depends on the effect. What I'm suggesting is there can be FTL effects that form a foundation for light-speed effects. These effects could be consensus-forming on light-speed expressions, functioning as part of a protocol for them.