A "for comment ballot" may not be the best time to get into details, but...

Some of the value sets still seem to be quite generic. For example, the BodyLocation value set seems to include a lot of locations that don't seem relevant to breast radiology reporting. Can these value set be pared down to relevant content (but still left as preferred or extensible to allow local expansion if necessary)?

As well, the Display values in many code systems (and downstream value sets) seem much wordier than expected and probably shouldn't just be a repeat of the Definition (this applies to most of the value sets). Several of these also contain typographical errors or inconsistent content. For example review DensityFocalAsymmetry, gunshotWound, DiagnosticTechniqueVS, FatContaining and HighDensity.