no liquid has uniform density (except a Bose condensate?)
That points to a misconception about superfluids, and reminds me of some
1) Any solid (even a "perfect" crystal) has nonuniform density at any
nonzero temperature, because of thermal phonons.
By the same token, any fluid (even a superfluid) has nonuniform density
at any nonzero temperature.
2) A crystal is ordered with respect to /position/. That is: If you
tell me the position of a few atoms in the crystal, I can calculate
the positions of all the others. (This does not mean they all have
the same position, merely that they have predictable positions.)
In contrast, a superfluid is ordered with respect to /momentum/.
That is: If you tell me the momenta of a few atoms in the
superfluid, I can calculate the momenta of all the others. (This
does not mean they all have the same momentum, merely that they
have predictable momenta.)
This has a nice interpretation in phase space: A crystal "crystallizes
out" along the position axis, while a superfluid "crystallizes out"
along the momentum axis.
You can have superfluidity in porous media, where there is no hope
of long-range positional order.