[theothermkn]

Only in the most general terms, and from first principles: Composite structures will have a higher stiffness per unit mass, which will cause the fundamental frequencies to be higher both for the pure structural modes and the control surface interaction modes. It's therefore likely that you'd only begin to encounter these aeroelastic modes at higher speeds. In other words, if you take the driving frequency as something like the inverse of the time it takes for air to pass over the wing, that may match the metal wing more closely than the composite wing. Modeling that stuff in a wind tunnel is very tricky, and you used to end up with these not-at-all-realistic-looking models that, nonetheless, captured some aspect of the full-scale aircraft being modeled.

Don't die. Stay in the envelope. Flutter is only the quickest way to ruin your airframe and day, not the only one.