Hello Aown,
This is definitely possible. You'd need to write custom operations though to achieve it. We do something similar except we automap data coming from sql stored procedures out to csv text files. Regardless of the input columns and types the export gets created. Sounds like you'll have a user do the mapping so all of that would need to happen ahead of time and outside of Rhino Etl (since we're dealing with the UI at that point.) I'd see it going something like this
DynamicExcelOperation
* constructor pulls in the fields that the user specified and initializes an object that can read from excel docs
* I can't help on excel libraries though since I don't use any of them today
SqlBulkInsertOperation
* constructor pulls in the mapping fields just like excel
* initializes the mapping for the columns based on what the user had entered
Process.Execute()
Nathan Palmer