CIT relies on the CU_PERSON_DATA database which pulls data from multiple sources, in some instances, they have to rely on views of Peoplesoft used to feed Salesforce. Some columns in CU_PERSON_DATA (see synonym) view are clearly marked "WD_" but are not being used.
Table documentation : https://cf.integrations.cornell.edu/Oracle%20Schema%20Docs/BOOMIPRDA%20(BOOMISEP)/schemas/BOOMI_FOLIO/tables1.html
The actual updates come from this table after the compare between what Cornell has and what FOLIO has: https://cf.integrations.cornell.edu/Oracle%20Schema%20Docs/BOOMIPRDA%20(BOOMISEP)/schemas/BOOMI_FOLIO/tables/FOLIO_USERS_DELTAS.html
Some questions can be answered by examining the source for CUPerson: https://cf.integrations.cornell.edu/Oracle%20Schema%20Docs/BOOMIPRDA%20(BOOMISEP)/schemas/BOOMI_FOLIO/views/CUPERSON.html
The population of most of the reference tables is not included here. Basically, when investigating anomalies, they start with CUPerson and work backwards.