It seems like the approach of cooling the organism to -30C at 350MPa, and then raising pressure further to ~600Mps to freeze it could actually solve that. As far as I understand, the speed of diffusion in water it far slower that the speed of sound (speed of sound at 25C is 1497 m/s, while diffusion coefficient for protons at 25C is 9.31e-5 cm^2/s, which corresponds to 1.4e-4 m/s - 8 orders of magnitude less), which is the speed of pressure gradient propagation. So if we use raising pressure as a way to initiate phase transition, it will occur nearly simultaneously everywhere, and the solutes won't have time to diffuse anywhere.

**ETA**: I just realized that since diffusion propagates according to inverse square law, while sound is linear, they should be compared to each other at the shortest distance possible. So I checked the time it takes for a proton to cover 0.1nm (hydrogen atom diameter) in water - 5.37e-13s, which gives us 186 m/s. It's far greater than the original number, but still an order of magnitude smaller than the speed of sound. And if we take 4nm (the thickness of a cell membrane) we have 8.59e-10s - only 4 m/s, so it decreases very quickly, and we're pretty much safe.

*10 points [-]