GAC binds organic compounds that include trace elements taken from the water column when they formed. Technically, that means that yes, GAC does remove some trace elements. IMO, the quantity removed is negligible and easily replaced with water changes. BTW, those trace elements bound in organic compounds are effectively unavailable to most other processes anyway, so it really doesn't matter that much.
How fast the GAC is expended depends on the quality of the GAC, the amount of organic material in the water column, the speed and method the water is exposed to the GAC, and to some extent, the bacterial activity in/on the GAC itself.
It is true that the organics trapped in the GAC are not exported and eventually break down. This allows nitrate, phosphate, and other byproducts of the biological process to be released back into the water column. However, when properly applied and maintained, GAC isn't a big nitrate or phosphate source. To not use GAC because it is a potential phosphate source would not be reasonable.
Some systems can benefit from GAC on a 24/7 basis. Soft coral systems need to regulate chemical warfare toxins, and heavily loaded system may need it just to keep the water from yellowing. If I were using GAC 24/7, I'd rather use a little changed frequently, rather than a lot changed more infrequently.
I like using a small reactor to ensure there is as little channeling as possible to increase efficiency and so I don't have to use a mesh bag which usually clogs long before the GAC itself. I use an old Phosban reactor and a bunch of sponges to fill the extra space. I'm lazy. The reactor makes it easy. I can rinse the sponges and change the GAC in a few minutes if I want, or just pull it out and dump the GAC if I'm not going to use it for a while.
I'm also a KISS fan but my interpretation of the acronym is a little different,. i prefer "Keep It Simple and don't be Stupid".