Hand-written org documentation goes stale the day it is finished, and nobody is funded to keep it current. Data Dictionary rebuilds the picture from the repo instead, so the value comes from never having to do that work by hand again. This page is a qualitative view of where that helps, not a spreadsheet of invented savings.
A traditional data dictionary is a deliverable: someone writes it, it is accurate for a week, then it drifts away from the org. Data Dictionary treats the same information as a build output, generated from the metadata already in your repo.
Custom objects, fields, automation and integrations
Functional testing, integration testing, UAT
Deployment, hypercare and ongoing support
Enhancements, scaling and new integrations