Cell-level lineage is the difference between explanation and proof
Column-level lineage shows where data may have moved. A reported figure needs a traceable path for the value itself.
Lineage is often treated as complete once tables and columns are traced. For many questions that is enough. For a reported risk figure, an IFRS 17 disclosure line or a Solvency II technical provision, the question is narrower: where did this value come from, which transformations touched it, and which rule applied?
Where lineage usually stops
In most estates lineage stops where the application code begins. The transformations that produce the reported figure sit in repositories the data governance estate cannot reach, and the final adjustment often sits in a spreadsheet. A lineage diagram can show where data may have moved. It cannot show what happened to the value in question.
Lineage into the code
xflow connects directly to the code repositories and brings the transformations into the same view as the process and the data. At a global systemically important bank, that closed a gap where lineage stopped at application code, and produced end-to-end lineage from policy to process to code to data.
Because each transformation is documented in enough detail to run, a value can be traced cell by cell and real data can be put through the process to prove it. That is the difference between explaining a figure and proving it.