Measure benchmark overhead on both Linux and macOS and fix any bottlenecks if excessive

[hassila]

Should validate what the overhead of the time not spent in the actual benchmark code is (capturing OS, malloc statistics, sampling) and see what the magnitude is - and optimise if needed - there are some cheap wins (especially on Linux, where we could cache some open files etc) that could be done if needed. Need to put in probes and measure overhead as first step (i.e. total wall clock runtime vs. measured wall clock runtime delta).

[hassila]

Several things addressed and basic measurements added as an empty benchmark that can be run (then use time to calculate overhead for e.g. a 100ms run). Caching of open files on Linux should probably eventually be added, but let's open a new issue if measurements prove there's an issue (simple to test with the empty benchmarks).