My guess is that it's not the strength of the fasteners that's being thought of, but rather the strength of the aluminum plates (also, the holes aren't tapped but rather simply drilled), and/or the compatibility of the metals involved.
My feeling was that if there was any hint of unfairness in either plate or any of the holes (not that there was anything obvious), that it would help to distribute the load to add sealant. While the fasteners should be plenty strong, what about any point loads? Maybe a non-issue, but since 4200 is removable, and since removal is not a planned event anyway (in other words, I want to be able to do it, but it doesn't have to be super convenient), I went with an all-over bedding of 4200 (less adhesive than 5200). In reality now, the fasteners are probably just backing up the adhesive of the sealant.
Another reason: In the case of stainless fasteners in a sandwich of aluminum plates that may sit in saltwater, I think the sealant would provide some isolation and help to prevent galvanic corrosion. Perhaps in that case Tef-Gel would be even better in the actual fastener area. (I actually used aluminum fasteners because their specs still looked plenty strong, and I didn't want to introduce steel into the aluminum sandwich.)
There's probably no way to know whether a given installation is overkill unless one tests to destruction - at least I don't know of a way to calculate it. I did select the fasteners and caulk type with knowledge of their specs, but there was a certain amount of "gut engineering" as well. I didn't want to be worrying about it, and the way I installed it felt comfortable to me.
Sunbeam :hot