First, it is assumed that the incident n is captured by the target nucleus and that an excited nucleus results with all populations of nucleon energies and angular momenta
populated with equal a-priori probability within constraints of energy and angular momentum conservation. In the HF approach(Hauser-Feshbach), this means that the n entry channel must consider the angular momenta (partial waves ) which the incident particle may bring in (usually from an optical model), the spin of the target nucleus, and the results of the coupling of incident spins with target a.m.The incident particle spin must also be coupled (e.g. 0.5 for a neutron). For the simpler Weisskopf-Ewing model-intended for low energy reactions where angular momentum constraints were less important, only the excitation of the projectile+target=compound nucleus is relevant;spin /a.m. is ignored as a constraint.
Now the statistical aspect: it is assumed that every energy/a.m (angular momentum) conserving final state will be populated with equal a-priori probability. The final state
consists of a product of the momentum cells of each 'p' of emitted particle,times its
intrinsic spin statistical factor(e.g. 2s+1 for a neutron=2*0.5+1)), times the number of final states left in the residual nucleus at its energy conserving final excitation based
on the p (and therefore energy and binding energy of the emitted particle). In W-E theory,only the energy dependence of final states in a given energy range U-U+DU considered, while in the more rigorous H-F approach,the outgoing particle/cluster has a limit as to partial waves it may remove according to energy and
residual nucleus. Additionally, the residual nucleus has a density of states dependent not only on its excitation, but in this more rigorous H-F approach,angular momentum (A.M). So to get the statistical probability of the final state, we need the density of final nucleus energy and angular momenta (e.g. 'spin dependent level density)-times the number of momentum cells for the outgoing particle.
Perhaps the answer to the question is that the cross sections are calculated assuming that every final state in the emission process is populated with equal a-priori probability; Weisskopf/Ewing calculated stats based on energy (momentum cells )only, Feshbach and Hauser added the consideration of angular momentum conservation to the calculation,so the final state densities have an energy and A.M. dependence.
Next are questions as to how the nuclear pollster calculates her statistics; biggest part is the level density- density of excited states in the residual nucleus- this is the tail that wags the dog. Other parameters are important, but less so.
But this old chemist rambles on- I hope that I have helped with an answer- tho I think that you knew all this-if so, my apologies. best wishes,marshall blann