Make comparison benchmark code for external compiler configurations clearer

[willzeng]

Right now this code is a bit hidden: https://github.com/unitaryfund/ucc/blob/main/benchmarks/scripts/common.py#L89-L110

This could
(1) be refactored to make it easier for contributors to suggest updates to the defaults that we use and
(2) be included in the docs so that contributors can easily find what defaults we mean when we mention qiskit and pytket on the benchmark plots

[bachase]

I'll focus on option 2 for as part of the next milestone, but will consider a refactored version as part of #235