What is the intrinsic scatter of the radial acceleration relation?
In plain words
If the universal curve has no spread at all beyond measurement errors, that would favor modified gravity. If it has a real spread that depends on galaxy history, that would favor dark matter.
Precise statement
Measure the intrinsic scatter (in dex) of $\log g_{\mathrm{obs}}$ at fixed $\log g_{\mathrm{bar}}$, after marginalizing over distance, inclination and stellar mass-to-light ratio, for rotation curves and for weak lensing out to $\sim 1\ \mathrm{Mpc}$. MOND predicts zero scatter and no residual dependence on galaxy properties; LCDM predicts nonzero scatter tied to halo concentration and formation history. An answer is the scatter with uncertainty below $0.02\ \mathrm{dex}$ and the residual correlations.
What would settle it
A homogeneous rotation-curve sample with independent distances plus stacked galaxy-galaxy weak lensing at low acceleration.
Status in the literature
Unverified note
2024 weak-lensing analyses reported the relation and flat implied rotation curves extending to several hundred kpc (moderate confidence).