Diffusion coefficient and halo size from secondary-to-primary ratios
In plain words
Some cosmic-ray nuclei, such as boron, are made when heavier nuclei hit interstellar gas, so their abundance tells how long cosmic rays wander. Converting that into how fast they spread and how thick the Galaxy's cosmic-ray halo is remains uncertain.
Precise statement
From $\mathrm{B}/\mathrm{C}$, $\mathrm{Li}/\mathrm{C}$, $\mathrm{Be}/\mathrm{C}$ and the radioactive clock $\mathrm{10Be}/\mathrm{9Be}$, determine the diffusion coefficient $D(R) = D_0 (R/1\ \mathrm{GV})^{\delta}$ ($\delta \sim 0.4 \text{ to } 0.5$) and the magnetic halo half-height $H$ to 20 percent, and decide whether diffusion is spatially uniform or strongly suppressed near sources, as TeV halos suggest.
What would settle it
Isotope-resolved 10Be/9Be measurements above a few GeV per nucleon combined with improved fragmentation cross sections.