I agree, those graphs were pretty confusing. As far as I can tell, they are not showing the average time each operation takes (ie average elapsed wall time), they must be taking the length of the test and dividing by the number of operations. It would be much easier to parse if they showed operations/second instead.

In other words, 9 concurrent pregnant mothers would show 1 month/baby on those charts.

If I have that right, their comment makes more sense. After 20 concurrent requests, the overall throughput of operations/second does not increase with additional concurrency.