Add conversion-table for RPM to latency

[cpaasch]

Tol ease understanding, we should have a conversion-table that would make it more intuitive to the reader to understand what RPM really means. E.g.,:

Latency (ms)	Responsiveness (RPM)
2	30000
5	12000
10	6000
20	3000
50	1200
100	600

(suggested by Michael R. Davis)

[moeller0]

How about to (optionally?) also reporting the delay measurements that the RPM was calculated from as output of the tool?
I would appreciate if such a measurement would also be performed without saturating load so one could see how far one's link "degrades" under load?

[moeller0]

So one thing that the RPM measure currently ignores is that it is an aggregate out of multiple data points, so it would make a ton of sense to also report some measure of the distribution of these measures (say to make things easy the standard deviation*), given the aim of not confusing the user, there can well be guarded behind a sufficiently scary command-line keyword (goresponsiveness reports the individual RTTs with the --debug switch, but it also includes a lot of additional distracting information).

*) Given the skewness of RTT distributions standard deviation is not the best statistic here, but one that can be computed easily and that is better than nothing...