You could be saving at least 20% of the time spent not having to worry about a datamodel.
You are not "not worrying about a datamodel" – you are just ignoring the data model that exists. And that overhead, if it exists at all, is far less than "20%".
You are not "not worrying about a datamodel" – you are just ignoring the data model that exists. And that overhead, if it exists at all, is far less than "20%".