The description of maturity levels is focused on resource exchange. These maturity levels are also applied to determine "maturity" of terminology service operations (and other operations?) but they do not describe testing that makes sense for an operation. Instead of "tested and exchanged" I would assume for an operation it should be tested with acceptable error messages and content exchange. There needs to be more documentation of what the "5 independent production systems in more than one country" have done to document that the operations have been tested.