In direct detection experiments, interpreting an observed rate as a limit
on (or detection of) dark matter requires assumptions about the type of
interaction. Typically, this breaks down as elastic nuclear scattering
(for energies >1 keV) or electron scattering (for energies 1-50 eV). The
first challenge to this conventional way of thinking came in the form of
the Migdal effect, wherein a sub-threshold nuclear recoil produces some
amount of detectable ionization due to coupling between the scattered
nucleus and atomic system. In semiconductors, this effect should be
amplified by inelastic collective effects and may differ substantially
from calculations using single particle wavefunctions. I present a
follow-up analysis of existing experimental excess rates and discuss
consistency with these scattering models.