Review choice of FlowDB config parameters to benchmark

[jc-harrison]

The full benchmarks suite already takes ~20 minutes to run (dependent on the machine where the benchmarks are run, obviously), and this will continue to grow as we add benchmarks for more query types.

On the other hand, we currently only run the benchmarks for one set of FlowDB config parameters, which gives us no indication of how performance scales with the size of the dataset. Information of this kind would be valuable for informing us of the current capabilities and limits of FlowKit.

We should review which sets of config parameters would be most valuable to benchmark, and how to do this (e.g. we could keep a daily run for one config, to track changes over time, and set up a second, less frequent, run to benchmark a larger parameter space).