Also if there are large generated files (the SQL parser is generated from a grammar) or if one project is vendoring dependencies it can artificially skew the li

Also if there are large generated files (the SQL parser is generated from a grammar) or if one project is vendoring dependencies it can artificially skew the line count. It's not a meaningful measure in itself without digging deeper.