I like the idea of parking bad data in a DQ database and tracking it's status each load until it is fixed in the source system. (Building a Data Warehouse With Examples in SQL Server)
I don't know how this would go when you are dealing with a relational data warehouse that has foreign keys. Do you also need to identify any children of the bad record and park them too?
I'd like to make this work, as it seems to be pretty comprehensive, but it has to work in a relational structure for us.