Inclined Fire, part #3.1: NDS, or the fall of the favorite

I started with NDS method first because their other methodologies seem to be simple,
straightforward, and highly applicable. I consider their sniper/DMR training and Perotti's
book on sniping top of the line.

In that book, "From 1 to 1,000", inclined fire is covered by a couple of paragraphs that say approximately the following:

1. No correction within 160 MIL (≃9.1º).
2. For larger inclinations
2.1. Calculate difference in altitude between shooter and target, either using a map or by multiplying the distance by sine of inclination angle;
2.2. Divide by 3
2.3. Subtract this number from linear distance to the target, resulting in "equivalent horizontal distance" (EHD).
2.4. Take adjustments as if shooting horizontally at EHD.

Item 1 is clear. Item 2 can be rephrased as EHD = D - D * sin(α) / 3, where D is distance to target and α is absolute inclination.

JBM and LO-Calc gave surprising results (I did not believe it at first, had to check twice):

GP11 vertical miss distance, cm

Looks pretty much useless. Even simply fixing sights at 350 m works better.

Accuracy: bad
Domain: up to ≃400-500 m
Complexity: moderate -- 1 table lookup, 2 mathematical operations
Pros: none
Cons: inelegant and useless

How this method made its way into otherwise excellent book is beyond my comprehension.
Looks like nobody's perfect.