In designing a human-computer interface (interface for short) for a complex work domain, the first question to be answered is what information should be presented on an interface media. The simplest answer is: it depends on tasks to be performed by the human operator. In the past two decades, two methodologies have been developed with a purpose to answer this question, namely Abstraction Hierarchy (AH) and Multi-Level Flow Modeling (MFM). In this paper we compare AH and MFM for the purpose of understanding whether they have provided a satisfactory answer. The result of the comparison concludes: (1) the two are complementary in terms of their functions, and (2) neither of them has provided a satisfactory answer to the foregoing question, and together they have not provided a satisfactory answer either. Subsequently, this paper outlines the criteria for a more satisfactory answer but the methodology to satisfy these criteria is considered as a future work.