The accepted solution of freezing one of the liquids arguably goes against the rule of 'not changing the molecular make-up of either substance'.

So a comparable solution would be:

1. Pour in the water 2. Pour in the milk. 3. Evaporate the combined liquid, removing all the water content and leaving only the milk proteins. (There are 3 or 4 distinct proteins in the milk of most species, all of which are distinct molecules, so technically this does not alter the substance on a molecular level. If changing the state from liquid to frozen solid is acceptable, surely this is too) 4. Collect the distilled water in another container (if you have to separate the liquids later then this must be allowable too) 5. Return half of the distilled water to the milk proteins to leave the original quantities of both milk and water.

The accepted solution of freezing one of the liquids arguably goes against the rule of 'not changing the molecular make-up of either substance'.

So a comparable solution would be:

1. Pour in the water 2. Pour in the milk. 3. Evaporate the combined liquid, removing all the water content and leaving only the milk proteins. (There are 3 or 4 distinct proteins in the milk of most species, all of which are distinct molecules, so technically this does not alter the substance on a molecular level. If changing the state from liquid to frozen solid is acceptable then changing it to a gas must be too) 4. Collect the distilled water in another container  5. Return half of the distilled water to the milk proteins to leave the original quantities of both milk and water.