Add test data for new septop gathering

[IAlibay]

We now have a specific Analysis unit for SepTop, which we use for gathering results.
Similar to AFE Protocols, this requires CLI gathering to have different behaviour (see #1937 ).

The code bits to address this were added in #1937, but we need to add new test files.

[IAlibay]

@atravitz do you know how the SepTop CLI gather test data was generated?

[atravitz]

I'm actually not sure, since the ligand names don't seem to match between the setup example and the analysis example.

@hannahbaumann @mikemhenry - do either of you know the source?

[hannahbaumann]

These are results from 4 Hif2a edges:

1 - 25
1 - 7a
1 - 7b
7a - 7b

[IAlibay]

After offline discussions with @hannahbaumann - we will regenerate the results from new inputs in openfe-benchmarks.

To avoid blocking the chain of PRs, this data will be added in a separate follow-up PR to #1937.