Add more detail on gromacs benzene dataset descr

[dotsdl]

The descr.txt for the Gromacs benzene dataset is a bit of a stub, and could use more detail, such as on:

• system size
• alchemical pathway
• temperature
• and anything else important for usage

Even a little ASCII diagram showing the alchemical pathway would help.

[dotsdl]

@ianmkenney would you mind taking this on, since this is your dataset?

[dotsdl]

Addressed by 90b07dd.

[dotsdl]

@ianmkenney can you please add the expected solvation free energy for benzene in water in units of kT, with a reference, to the DESCR? This would give us a "right answer" to directly compare to. Could even make this an attribute of the data Bunch, too.

[orbeckst]

We didn't include a "right" answer for the Amber data #5 either. Philosophically, that's also a problem because this will depend on the method that is used to calculate the "right" answer from the data. For the tests, this is the reference value and it should live in the test case. Whoever writes the test will determine what the correct reference value for the method and parameters is going to be.

I think it is enough to include an experimental value, if one is available, to get a sense where the computed value should come down.