Protocols have grown larger and more complex with the advent of computer and communication technologies. As a result, the task of conformance testing of protocol implementation has also become more complex. The study of design for testability (DFT) is a research area in which researchers investigate design principles that will help to overcome the ever increasing complexity of testing distributed systems. Testability metrics are essential for evaluating and comparing designs. In a previous paper, we introduce a new metric for testability of communication protocols, based on the detection probability of a default. We demonstrate the usefulness of the metric for identifying faults there are more difficult to detect. In this paper, we present two approaches for improved testing of a protocol implementation once those faults that are difficult to detect are identified.