Review all included benchmark systems to ensure they are compatible with current best practices

[jchodera]

We need to review all current protein-ligand benchmark systems (and their associated crystallographic source data) to ensure they are consistent with the current best practices for inclusion in high-quality free energy benchmark sets. Some of the systems have issues that may preclude their continued inclusion, even if they were included historically.

best practices paper: https://arxiv.org/abs/2105.06222 source for paper: https://github.com/openforcefield/FE-Benchmarks-Best-Practices

[dotsdl]

Can we put a checklist here for criteria that must pass to keep a benchmark system, derived from the best practices paper?