Ethnography as a software requirement elicitation method

I am looking to understand the use of Ethnography to illicit requirements in Software Engineering. Are there two good examples of when Ethnography may not be the ideal choice to illicit requirements?