OK Tom. Now I understand the situation. And I agree with you that the book answer is wrong. But I seem to vaguely recall that this issue has been debated here before (not recently). I do not recall whether the membership arrived at a consensus.
First I’ll comment on your first example (Post #7). The question was, what is the minimum lighting load. To get that answer, you look in 220, not in 215. Anything 215 has to say about sizing feeders will have nothing to do with the question of the amount of lighting load. The answer is 10,240, not 12,800.
Now suppose we were to say that all of the lighting load, and no other load, was going to be supplied by a single, separate, panel. Suppose we then asked what size the feeder to that panel had to be. Now you are into 215. Here we must pause and consider whether to use the 10,240 value or the 12,800 value. This is where the debate came in.
? Some members (myself included, I seem to recall) believed that you never add 25% to a load value that originated in the “watts per square foot table." The notion is that that table represents a lighting allowance, not an actual value of lighting load. If you instead arrived at a lighting load value by counting the actual fixtures that were installed, then that would be a “continuous lighting load,” and the 125% factor would be applied.
? Other members believed that lighting is continuous in most real-life applications. That is, lights are going to be on more than three hours in a row in most offices, warehouses, theaters, banks, and other buildings. They conclude that the “watts per square foot” value of lighting load needs to be treated as a continuous load, and they apply the 125% factor.
So depending on which side you take in this debate, the answer to my proposed question would be one of the following two values:
? You size the feeder for a minimum ampacity corresponding to 10,240 VA, based on a lighting load of 10,240 VA, or
? You size the feeder for a minimum ampacity corresponding to 12,800 VA, based on 125% of a lighting load of 10,240 VA.
But whichever of these you choose, the lighting load is still 10,240 VA. I say again, the book is wrong.