I think it would be interesting to add artificial benchmarks, for example, to compare how methods cope with collections of sequential graphs (permutations). Maybe the results will look better as a graph which shows how different methods work for, say, 5 events and up possible sequential graphs (5! = 120 permutations of 5 events, then also 5 * 4! = 120 permutations of 4 events, 10 * 3! = 60 permutations of 3 events, etc.).
I think it would be interesting to add artificial benchmarks, for example, to compare how methods cope with collections of sequential graphs (permutations). Maybe the results will look better as a graph which shows how different methods work for, say, 5 events and up possible sequential graphs (5! = 120 permutations of 5 events, then also 5 * 4! = 120 permutations of 4 events, 10 * 3! = 60 permutations of 3 events, etc.).