And even if they all run at the same time, some of them might be running at less than their full load value. OK, this explains that the loads can be diverse. But it does not provide a definition of the two word phrase, "load diversity," as that phrase is used in the Informational Note cited in post #1. If you were interested in hiring me, as a PE, to perform an ampacity calculation that takes advantage of the NEC's method for applying "load diversity," I would decline. I absolutely do not know what that phrase means.
You might suggest that it means, as an example, that there are 12 fans powered by the same circuit, but they are interlocked such that only 6 can run at the same time. But that is not "load diversity," but rather a situation that involves "noncoincident loads," as discussed in 220.60.
You might suggest that it means, as an example, that an apartment building has 20 units, each with its own range, and not all ranges are likely to be in use at the same time. Here again, that is not "load diversity," but rather a situation in which demand factors come into play.
The link provided in post 4 discusses demand factor and diversity factor. Demand factor is covered well in the NEC. Diversity factor is not addressed at all. It is not the same as load diversity, and would not help in any service load or ampacity calculation.
Bottom line: I wish the NEC did not offer a calculation process that employs "load diversity," without taking the trouble to define the phrase. It is not a phrase that has a well understood meaning in the electrical engineering field.