Perhaps I should explain in more detail why phosphate bound to a CaCO3 surface is not really a good analogy, IMO.
Phosphate does bind to calcium carbonate surfaces (such as live rock and sand) in much the same way that ammonia and other ions (potassium, etc) bind to a zeolite. Both types of binding are reversible, more binding happens at higher ion concentrations, and may be impacted by factors like pH.
In the case of phosphate bound to CaCO3, that will happen any time phosphate becomes elevated in the aquarium. The surface bound phosphate then acts as a reservoir, releasing that phosphate back into the water if the concentration in the water drops. Additionally, phosphate can be released from CaCO3 surfaces by ph changes. The maximum binding happens roughly around pH 8.4, and either higher or lower pH will release some of it.
Algae that have become attached to the rock surface can locally impact the pH, by using up CO2 during photosynthesis, or generating it at night by respiration. Some species may also be able to impact pH in other ways as well. If the algae substantially changes the local pH between itself and the underlying live rock, it might cause phosphate that is already bound to be released, and become available to it. So in a sense, the rock can act as a source of phosphate for the algae when there is not enough in the water column itself.
Note, however, that that is a one way process. It is not an ongoing effect. Once the rock is depleted of surface bound phosphate, there is no more to be had (without dissolving away more rock with very low pH, which may happen, but is itself a one way process). Specifically, this process does not work by using the CaCO3 as an ongoing collector of phosphate for the algae. A fresh CaCO3 surface with no phosphate is going to be scavenging phosphate form the water, and that will take it away from algae, not provide it. Any time that phosphate is actually adding to the rock, the algae are not getting it and the local concentration near the algae is lower, reducing the bioavailability.
So for phosphate on CaCO3 surface, once any surface reservoir is depleted, the algae does not benefit from this process. This is easily verified in many ways, not least of which is the various processes used to strip phosphate from CaCO3 surfaces (rock cooking, acid washing, etc) that make it much less likely problem algae will reappear.
So, how is this different from ammonia on a zeolite? The second part is not. If you start with a zeolite with no ammonia on it, it is exactly analogous to phosphate on CaCO3: the algae or bacteria do not benefit from having nutrients bound to it.
If, somehow, you bound ammonia first then added bacteria, yes that could possibly benefit bacteria, just as phosphate bound to rock can help algae. The bacteria may be able to alter the local environment enough to reduce ammonia binding (not sure exactly how that would happen, but it is at least theoretically possible). But once it is done, it is not an ongoing process, just as with phosphate.
So the idea that the zeolite collects ammonia from the water and serves it up to the bacteria is flawed. It may happen once initially (if you bind ammonia before bacteria colonize the surface), but once the effect happens, the binding sites are available again, taking ammonia that the bacteria would otherwise potentially get (or it does nothing if the bacteria have permanently altered the zeolite surface in a way preventing ammonia from binding.
Anyway, I hope this makes sense to folks and helps explain why I think the ammonia release hypothesis is not valid.
