PaulP: "It's not about the predictions, but about the physics."
More broadly, ENSO prediction geophysics are much the same physics as weather prediction. There is a lot human life and assets riding on the best physics-based predictive models. Technical forecasters look at multiple physical models to make the best predictions.
You seem also to speak to the concern your model might not be predictive "by chance", yet still be physically correct. But what about being studiously uninterested in how much Lunisolar noise is directly picked up by ENSO sensor networks, to contaminate the data? Its not about that either?
An alternative Topological Insulator identification of the Equatorial Kelvin Wave is the major insulated bulks being Doldrum zones North and South of the narrow waveguide, and the wave itself being rather ballistic, not so much statically insulated. Once again, its the ontological Elephant Problem. At least we can predict some of the physics easily- Kelvin-Helmholtz Instabilities at the interface surfaces of the topological bulks.