Standard cosmography parameterizes cosmic expansion without assuming a model for dark energy or dark matter. It's agnostic but limited — it can't capture physics like dark matter decay, where a parent particle splits into massive and massless daughters.
The authors (arXiv:2603.23247) build a semi-cosmographic framework: Padé-rational parameterization of the luminosity distance (model-agnostic for dark energy) combined with a specific two-body decay model for dark matter (phenomenologically motivated). No dark energy model is assumed, but the dark matter sector has structure.
Fitting to DESI DR2 BAO data and forecasted 21cm intensity-mapping observations, they reconstruct both the massive daughter's equation of state and the inferred dark energy equation of state. The dark energy EOS is not assumed — it falls out of the framework as a residual after the dark matter decay is accounted for.
The through-claim: what we call “dark energy” may be partly an artifact of assuming dark matter is stable. If dark matter decays, the daughter particles modify the expansion history in ways that would be misattributed to dark energy under ΛCDM. The semi-cosmographic approach separates the two: it gives dark matter decay its own degrees of freedom and lets the data decide how much of the expansion anomaly is dark energy and how much is decaying dark matter.