Is the underscreening scaling exponent 3 universal across electrolytes?

Define $n$ by $\lambda_S/\lambda_D \sim (a/\lambda_D)^n$ in the concentrated regime $a/\lambda_D >\sim 1$. Determine whether $n$ takes a single value across aqueous alkali halides, multivalent salts and ionic liquids, and whether that value is 3 as surface-force data suggest or the $n \sim 1.5$ range of bulk liquid-state theory. An answer is a measured or computed $n$ with error bars for at least three chemically distinct classes.

Systematic measurements of $\lambda_{S}$ across salts and solvents with at least two independent techniques, fitted over a decade or more of $a/\lambda_{D}$.