A potential sink that has been described by some hobbyists is the precipitation of magnesium by limewater (kalkwasser). Both magnesium hydroxide and magnesium carbonate have been suggested. I do not believe that either is an important process in most aquaria. Adding any high pH additive, including limewater, results in the transient formation of magnesium hydroxide. This material quickly redissolves on mixing such that the local pH drops below about 8.6.-9.0. Magnesium carbonate is a more complicated issue, as it is near its solubility limit in seawater and may quickly get coated with a less soluble magnesium calcite. These issues have been dealt with by
Bingman in much greater detail, and his conclusion is that neither of these precipitates is a likely sink for magnesium.
I would suggest that an alternative way that aquaria using only limewater might become deficient in magnesium over time is that the limewater is simply not delivering magnesium to the aquarium even though it is present in the solid lime. How and why this might happen was discussed above involving the precipitation of magnesium hydroxide in the limewater reservoir. This lack of addition coupled to the ongoing removal of magnesium in calcification could lead to deficiencies in magnesium. Such deficiencies have not become extensive in my aquarium, but it does not have an especially high rate of calcification, and perhaps water changes have eliminated the problem. In any case, those using only limewater (or other systems that do not deliver magnesium) may want to occasionally check magnesium.