Because of the fact that the sql contains heavily nested sub-queries an appropriate incremental control-logic might be placed n times. I suggest to start with an optimizing of the sql - for example by removing the nested queries and simplifying as much as possible.
And maybe you ends with n quite pure raw-data queries to Qlik and applying the wanted transformations there. In each case you need to understand exactly what the sql is doing to be able to apply anything on top of it.
My suggestion will be to create a stored procedure using the script which you have shared above into a table in SQL. Further use incremental load based on the progress_date column/Data field from the table to describe the QVD generation. this way your load will be quick and you will load only the new records as part of QVD creation. Happy to help you with optimization in SQL as well as Qlik Sense.
