You might be interested in what we have done with our ADF Add-On product, ChillETL.  It is completely metadata driven, and works with any ADF pipeline parameters.

I wholeheartedly agree that dynamic ETL is the way to go.  That is why I have developed ChillETL, which is a totally meta data driven Add-In for Azure Data Factory.

We developed ChillETL several years ago as a completely meta data driven way to use Azure Data Factory Pipelines.  I believe it goes well beyond the feature described here...