This paper deals with the testing capabilities of the modern methodology of science in the specific problem area of cosmology. I propose to monitor the responses of selected competing methodologies (Kuhn, Lakatos) on two questions: 1) To what extent are modern methodologies able to cope with the problem area as a descriptive task? and 2) to what extent are modern methodologies able to cope with the problem area as a prescriptive task? In both cases it appears that the tested methodologies find themselves in a quandary when trying to submit answers. The conclusion of the article attempts to show that this quandary remains, in a certain form. However, this does not mean that methodology should give up its fundamental role, ie. the role of being a heuristics in achieving knowledge