--maxrate and --minrate options

easy-cass-stress can either use fixed --rate or variable rate with --maxrlate / --maxwlat. This generally going to be good enough. However, there are going to be more complex cases where people will want to run multiple workloads against a single cluster.

If that happens now, the results could be a bit unpredictable. I'd like to be able to set a minimum and maximum throughput, which will make it so one workload doesn't completely back off while the other one picks up.

Start by tracing how easy-cass-stress currently parses --rate, --maxrlate, and --maxwlat and applies variable-rate behavior. Define how minimum and maximum throughput should interact across concurrent workloads, then verify that one workload cannot completely back off while another takes over.

Written by the indexing model from the issue text.