initial benchmarks to try verifying O(2^k * n) for graph coloring

[andersenwall]

initial benchmarks complete for constant 2 bags with n nodes. verified n linear for false returns on coloring tests problem found with n worse than exponential for true returns on coloring tests

[jpfairbanks]

@andersenwall, you have permission from @samuelsonric to replace the failing test with a check that the lengths are equal

[andersenwall]

@andersenwall, you have permission from @samuelsonric to replace the failing test with a check that the lengths are equal

resolved, test cases should all pass now