Minor issues with multi-softening

There are a couple of minor issues with multi-softening gravity that we'll track here:

• Dry runs don't work because of line 978 in parallel_io.c, which sets the gravity particle mass for the high resolution particles. If that's not there, it looks like PartType1 doesn't ever exist. We would probably have to read just a single gravity particle mass, or else fall back to the MassTable?
• The new parameters are not documented in parameter_example.yml