I've read through this old thread and here are some of my thoughts:
You might expect 12 times the concentration of hydrogen and hydroxide anions as your pH swings, meaning that many trace metals could quite literally break their existing hydrogen bonding (in the substrate - not likely in your case)
I'm not sure whether pH moves that markedly in the sediment as a result of CO2 on/off injection. Firstly, the diffusion between sediment and water column is rather slow. Secondly, in the sediment, pH is mainly influenced by factors in situ, primarily by microbial activity, and also by plants' roots. Thirdly, Ponnamperuma has studied pH of submerged (flooded) soils and demonstrated that they tend to approximate neutral pH, no matter if water column is acidic or basic.
I think activated carbon is one of the best suggestions I have ever read for dealing with trace metal toxicity
Again, I'm not sure. Couple of years ago I've studied this question quite extensively. Adsorption of transition metals on AC is indeed possible. Yet, AFAIK water processing plants do not use AC for that purpose, it's not sufficiently effective. Aluminum and iron (hydr)oxides are stronger adsorbents for these metals (albeit together with phosphates and silicates).
Of course, treated (coated) activated carbon may be a wonderful adsorbent for selected metals (or other species). The problem is price, and for us hobbyists - availability.
It is strange that you are having this problem on an inert substrate.
I think a sandy substrate is inert only in the beginning, a few weeks, perhaps. Dead roots, lysed microbes, their biofilms, plus dead organic matter settling from the water column quickly form a significant mass of detritus. Detritus is known to be quite strong adsorbent for various species. The more so if in interaction with iron oxides and clays.
My experiments suggest that one year old sandy substrate is a significant reservoir of
all nutrients. (I've attempted to document signs of complete lack of selected nutrients. I've made a 100 % WC and re-mineralized with the nutrient in question completely missing. I've waited a month, two, three, ... and no signs of deficiency appeared, while the plants kept on growing. Since then, I do not underestimate detritus as a source of nutrients.)
Plants have a cobalt (Co) requirement
Have they?
Just don't add the trace element solution on a regular basis, wait for the start of deficiency symptoms.
Microelement toxicity probably happens more often than we admit. It's quite difficult to diagnose.
I think that "complete micro-fertilizers" are very dangerous. Of all micros, iron is in want most often, by far. That's thanks to specific features of iron. Unfortunately, when signs of iron deficiency appear, many hobbyists dose
complete micros! Micros other than iron then accumulate.
Additionally, micros adsorb in the sediment. Therefore, no water change will dilute them. Then, if disturbed in some way - physico-chemical changes, but very likely re-planting or re-arranging hardware is enough - they may release in the water column in toxic amounts.
I think an interesting way of dosing micros is "fertilizing" with food-grade Chlorella or Spirulina powders. Sadly, both of them lack boron (boron is essential only for higher plants), but Spirulina definitely contains cobalt (it's an essential element for cyanobacteria). I've experimented with these fertilizers briefly, and it works. Plants grew for three months without any additional micros dosing (RO+DI water, no fish feeding etc.). The drawback was that I've got suspicion that they supported algae. More tests would be necessary to assess this question with more clarity.
In general, I'm of the opinion that nutrient deficiencies occur more often in more abundantly fertilized/mineralized tanks. I think the reason is that nutrients are not absolutely missing but their availability is hindered by high concentrations of other nutrients. Most of the time, I work with soft water and lean dosing, and I rarely observe signs of nutrient deficiencies.