To be fair, though, the Lawful or Chaotic character likely does consider either order or freedom to be an end, and is likely baffled by those who simply treat them as means toward an end. This goes back to an earlier exchange about treating Law and Chaos differently than Good and Evil. I'm more concerned with the latter than the former. But you have a point with respect to the former.
In the case of Evil, those means are also an end. Evil doesn't hurt, kill, or oppress for some utilitarian purpose or ulterior motive. It hurts, kills, and oppresses because it enjoys causing pain.
Some evil does this. Some evil causes this pain because it is indifferent to the pain it causes. The rules seem, from my reading, to acknowledge both a sadistic evil and an indifferent evil. The rules don't seem to support the idea that evil always equals sadism; it sometimes equals sadism.
The enjoyment of the torture is the reason the torture is happening.
Would you make that argument in today's world? I'm an annoying Canadian socialist and even I wouldn't go that far.
I would argue (and have argued) that if a character is torturing an NPC toward a utilitarian end (or helping them toward a utilitiaran end), those acts are not necessarily Good nor Evil. What makes a character Good or Evil is that their behavior has no other motivation.
But the rules as written contradict this statement. The DMG specifically says that a DM should not take notice of stated intentions when determining alignment and must only take notice of PCs' actions.
And how does this relate back to the question of which alignment has an intellectual edge?
My Coventry example relates back in the following way:
(a) intelligence appears to be a criterion for alignments that permit means-ends distinctions; the more sophisticated the means-ends distinctions, the narrower the range of alignments available.
(b) I brought up Coventry because the D&D alignment system poorly aggregates the data in any situation where a group is sacrificed for a greater good and produces essentially incoherent results
With respect to the things that the D&D alignment concerns itself, I think it does OK. Just don't expect it to model distinctions in the moral spectrum of the real world that it doesn't address and don't be suprised if it creates some strange bedfellows.
This term "strange bedfellows" I assume is what you mean by the fact that anyone with a complex relationship between their goal and strategy is returned as "neutral" under the system. This, returning to Coventry, is my point: I think there is something wrong with the system when it places radically dissimilar individuals in the "neutral" category simply because their goals and strategies exist in a sophisticated or complex relationship to one another.
But regardless, I think you'll also find geniuses and morons across the D&D alignment spectrum, if you want to think of that as a distinct spectrum.
While true, I would argue that D&D is predisposed to place geniuses in the "neutral" category, which is especially annoying because that's the same category it uses for creatures too stupid to have a morality.