Testing: Generate more robust testing graphs

[cmalinmayor]

• Manually create a wider range of graphs with more edge cases for testing all metrics. Organize these fixtures or functions better - we currently have some fixtures inside the TrackingGraph tests and some inside test_utils.py. We shouldimprove naming and organization as we manually generate a wider range of graphs.
• For the CTC, add more comparisons with values from the executable (manually or programatically).
• Look into synthetically generating graphs for CTC and/or for the rest of the tests.