The expansion of the steam across the leak or fitting is at constant enthlaply from a pure thermodynamic point of view.
However, there are complicating factors in the real world. Once the steam is in the atmosphere, there is heat transfer and vapor pressure considerations as it mixes with the 'cold' ambient air. That will cause some steam to condense into 'fog' as the air can't hold all of the steam in vapor form. Note, while I've seen some water dripping off the piping at a steam leak, I haven't seen a spray of water droplets coming out which is more what I'd expect IF a constant entrophy expansion (versus enthalphy) was occuring.
In the real world, the cloud of condensed steam simply mixes with more ambient air until the condensed water is all picked up again. This assumes there is enough ambient air (ie. not inside) and enough physical room before it hits the surrounding (ground, beam, vessel) where some of the condensed steam could 'stick' to the solid object (note the highly technical term 'stick', there must be a better term, can't think of it though). Plus, if the leak is small, heat transfer from the escaping steam through the 'fitting' to the surrounding piping/fitting and then to atmosphere might cause some condensation, that would cause the drips you sometimes see. Any significant steam leak I've seen from a valve packing or a fitting more resemblances a steam jet than a mixture of steam and water droplets.
I've got a really nice little steam program from Archon Engineering (highly recommended BTW) that I'll run some numbers on tonight to show what the outlet temperature temperature would be for a constant enthalphy versus constant entropy.
However, I can assure you it is an constant enthalphy expansion, this is the principle used in a throttling steam calorimeter to quantify steam quality or its initial temperature.
