Getting XBRL-based digital financial reporting to work effectively is not about solving one big "problem"; it is about solving lots of little problems. Think about the problem in terms of a "template".
Here is a template within a ZIP archive. (Here is the human-readable information.) That is what a software application must manage, the successful creation of that template by a business professional.
Now, the template could take the form of any of the patterns in this proof. You need ONLY support those patterns. This is the architecture. A rules engine watches over everything. Boundaries are well defined.
The templates might be for different reporting schemes: US GAAP, IFRS, IPSAS, MINI, FRF for SMEs, XASB. But every one of those templates would follow the same set of patterns in this proof.
An XBRL-based digital financial report is not one big thing; it is lots of little things. The full report can be thought of as a stream of templates. Each template must be sound and each template must interact with other templates.
Software must control the possible "impediments" to success. Here are those: Impediments 1, Impediments 2, Impediments 3. Those impediments are overcome using rules which prevent information anarchy.
This is how all of these examples work.
This video play list, Understanding the Financial Report Logical Model, explains the details.