These opt-in panels are merely profiled volunteers that do not have to live in the state. That makes such polls fairly worthless. Remember MOE is only a projection that if the poll is done exactly the same way again that it should come in within the MOE of the same number. Folks sometimes falsely believe that MOE is a guarantee of the goal of the poll rather than its methodology.
Good points
Since they included just landlines and the “opt-in” I assumed it had something to do with mobile phone responses.
But the biggest error in the poll was starting if off with “Do you approve of Mussolini?” “Do you approve of the KKK?”
Asking those questions last would have been alright.