WiredTiger test coverage can be measured by the wiredtiger-test-coverage Jenkins job. We should periodically, either manually or automatically, review this coverage and identify/fix areas that have poor coverage.

As a small related issue, the current job includes running programs in the examples/c directory, and it really should not. The examples have a different purpose, to demonstrate API calls, and don't necessarily check that the calls actually did what they promise to do. If we lose coverage by not including examples, we should add explicit tests.  We should also include test/csuite programs in our coverage testing, and use explicit configurations when running test/format to maximize our coverage.