ASTRO
In the literature: contested
Does the GeV excess follow the stellar bulge or a spherical halo?
In plain words
Pulsars would follow the elongated, box-shaped bulge of stars, while dark matter would make a round glow.
Precise statement
With interstellar-emission model uncertainty marginalized (families of cosmic-ray propagation models, template or nonparametric fits), determine whether the spatial distribution of the excess is better fit by a boxy or X-shaped stellar-bulge template or by a spherically symmetric squared generalized-NFW profile with inner slope $\gamma \sim 1.2$. Answer: a likelihood ratio or Bayes factor with systematic uncertainty that is stable across foreground models.
What would settle it
A morphology preference that holds across all foreground models consistent with the rest of the gamma-ray sky.