My latest project involves actions returning different views at runtime (simply put, the application has a number of different customers as users, and they want their own custom data entry form designs - these forms all share the same model object-graph, but their display of the model's data is radically different to each other)...
For example, a "Case" entity has multiple People associated with it. Some forms only display a single Person's fields, other forms support multiple Persons, and another doesn't contain Person information at all. Furthermore, the forms have differing levels of fields - many forms lack many fields that others have.
Because of this radically different behaviour, I believe I'll need a different model binder for each view design.
Can StackOverflow recommend a course of action to take in this case, or an alternative solution that keeps the code simple.