Benchmarks: more "real" and MILP benchmarks for Q3

[siddharth-krishna]

We're adding new benchmarks from the following energy models:

• PyPSA (to match the features seen in the GenX benchmarks): Benchmarks: new PyPSA benchmarks with missing features #91
• GenX (to get some "real" sized examples): Benchmarks: new GenX "real" benchmarks #87
• Temoa: Benchmarks: add Temoa benchmarks and instructions #89
• TIMES
  • Also in touch with: Evangelos, James Glynn
• Balmoral

For each new benchmark, let's ensure that we add:

• Metadata to benchmarks/<model-framework>/metadata.yaml
• The LP/MPS file to GCP buckets
• The GCP URL, name, and size to benchmarks/benchmark_config.yaml the metadata file
• Run python benchmarks/merge_metadata.py before opening a PR to ensure metadata is synced
• If possible, the configuration file and instructions on how to generate the benchmark to benchmarks/<model-framework>/. Large data files should not be checked into this repo, but hosted on another site and the URLs can be added to the benchmark generation instructions instead.
• Let's also try to maintain the convention that benchmark names are prefixed with the model framework's name, e.g. pyspa-... or temoa-...