When should an external rating be considered as a primary factor determining an internal rating assignment

Concerning the practical application of the RTS on materiality: When should an external rating be considered as a 'primary factor' determining an internal rating assignment, and by what measure is the relative importance of the different variables comprising an internal model for credit risk to be evaluated?

According to Annex I, part 2, section 1, paragraph 2(e) in the abovementioned RTS, the introduction or withdrawal of an external rating as a primary factor determining an internal rating assignment should be considered as a change requiring competent authorities' approval (i.e. ‘material’). However, it is not specified what is meant by 'primary factor' in this context. Neither is it explained how the relative importance of the different variables included in an internal model for credit risk is to be evaluated. For example, when an external rating is included as one of the risk drivers in a statistical scoring model, it is usually found to be a 'strong' variable (this can be evaluated by several different means, such as information value ordering or by comparing the Wald Chi-2 test statistics). Obviously, if the external rating turns out to be a very dominating variable (say > 90% of the total weight in the model) it should be considered primary. And if it turns out to be the least important variable it should perhaps not be considered primary. But what if the external rating is found to be the second or third most important variable in the model?