Hi mpat,
the option 1 (assume free water present) in Pro de Properties produces conservative values which are in most cases good for preliminary evaluations, with this option the software assumes there is water in liquid phase even when water has not been included in mixture.
As you know free water is not required for hydrates formation.
The other option which I utilize quite frequently is to force a VLE calc. for water phase, this allows to evaluate if liquid phase is present and generally produces less conservative values.
The last option (solve as multiphase equilibria including solids as ice) should give accurate values but I prefer conservative values so I select options 1 or 2
As far as I understand with multiphase option the procedure keeps in account the different fugacities of water and other components when in solid phase, these are different from fugacities of liquid phases hence some differences in resulting values.