Is it Academically Acceptable to Make Your Own Qualitative Research Method Based on Existing Ones?

Please consider the following scenario:

Suppose that researcher X is a qualitative purist who thinks that contemporary qualitative research has a tendency to lean towards quantitative methods. Suppose, too, that he decides to use observational methods and discourse analysis to study a certain political phenomena. However, since these two methods are increasingly becoming dependent on statistical and other formal methods, he decides to create a new variant to fit his purist research orientation. Is it academically acceptable for him to do so? Do you know of any published articles or books that thoroughly discuss this issue?