In Enterprise Architecture (EA) model analysis activities, it is critical to make early statements and diagnosis from a high level of abstraction. Currently, these tasks are difficult to perform, and they require both the involvement of experts and the elaboration of specialized artifacts. Further-more, the complexity of the tasks increases as models become bigger, more detailed, and start covering more domains. In other contexts, it has been noticed that total / holistic / unfiltered visualizations may give insight about the models, providing analysts a starting point for exploration and general pattern discovery. In this paper, we offer an evaluation framework to assess the strengths and weaknesses of visualization tools to support EA analysis activities. The framework is based on a set of 14 requirements which are either visualization-related or specific to EA analysis, and which where harvested from experimentation and from a survey of current EA and visualization tools. The requirements were defined around the design principle of cognitive effectiveness, and they favor a broader use of visual languages in order to enhance the semantics of EA visualizations.