Don't give up yet. There is more shared here than you are allowing.
You are farther along than me in showing your ENSO model results, after so many years head-start. Anyone needs the sort of time it took you to be ready. Give me time to satisfy your requests for even more, until you can fairly concede some value. I have at least posed my starting hypotheses.
Chaos science does not "punt", but meets its dynamics head on. Weather science is a brilliant success story. Sensor noise is no blind spot not to be spoken of. If your model is merely memorizing the noisy data, to seemingly recreate it accurately, then predicting the next few years will show gross mismatch. If your model is true, actual and predicted wave forms will match closely. If the Lunisolar Noise Hypothesis is true, your model will well predict that signal.
We'll see if a combined model, with better sensor-noise filtering, does better. Fortunately its not all on me. We will accept anyone's multi-physics model and better data validated by superior predictive power. In particular, solar radiation, seasonal variance, and geological harmonics in a model seem essential to best predictive performance, apart from any Lunisolar components.
Demanding immediate third-party validations here, while invoking 2280 as "even better", is wide of the mark on both ends. I devote to diligent study on a reasonable timescale. Thanks for any encouragement.