Due to the fact that we receive data in a loop through the Rest connector, we get large empty files. Also, such metadata (hundreds or thousands of lines) is of no use.
If you run one Transform application with multiples data transformation steps, then the last saved QVD will contain all the steps for all files processed in this Transform. That is, every tab in the Qlik Sense script prepares 1 QVD file from many other sources (qvd, database, Excel, etc.). On each tab, it would be logical to reset the cache of data lineage.