Predictive time-dependent density-functional dynamics of fission yields
In plain words
Time-dependent density-functional theory, a method that follows the motion of all nucleons through an average field they create together, can simulate one fission event but gives only average outcomes. Turning it into a method that predicts full distributions of fragment masses and energies without tuning to fission data is still unsolved.
Precise statement
Extend time-dependent density-functional theory with pairing (TDHFB, TDSLDA) by a beyond-mean-field treatment of fluctuations (time-dependent generator coordinate method, stochastic mean field, or time-dependent random phase approximation) so that it predicts the mass and charge yields $Y(A)$, $Y(Z)$, the total kinetic energy distribution $\mathrm{TKE}(A)$ and their widths for $\mathrm{235U}(n_{\mathrm{th}},f)$ and $\mathrm{252Cf}(\mathrm{sf})$ within experimental uncertainty. The functional must be fixed by non-fission data only, and the answer is a demonstration of such predictions.
What would settle it
Published yields and kinetic-energy distributions from a fluctuation-including microscopic method, with no fission-data fit, agreeing with evaluated data for several actinides and predicting a measured but unseen case.