Can you use the number of test cases as a metric? In this blog post, James Christie discusses this topic and the usefulness of counting test cases.
His point of view is that the number of test cases and their use as a metric of progress for the software testing effort are not meaningful. What is important is how the tests run are important to validate the software under test. You can have a lot of small test cases that give limited information and a complex test case which is difficult to run but provides valuable evidence that the application is running correctly. In his experience this is however sometimes difficult to explain to management who just want reports to “manage the stakeholders”. Whether the metrics are really meaningful to assess the quality of the software is less important.
Read the complete blog post on http://clarotesting.wordpress.com/2010/07/21/but-how-many-test-cases/