You can read the relevant sections (at least δε and ἀλλα) in
http://www.ntdiscourse.org/docs/Discour ... sample.pdf. Better yet, buy the whole book and read it
To answer your question: the textbook is misleading, if you represent it correctly.
ἀλλα is quite much an "explicit" 'but', like the English equivalent 'but' - although this, too, is simplification. In a way your textbook is correct. To quote Runge, "Heckert is able to reach more specific conclusions than “adversative” in his description of
ἀλλὰ. He describes ἀλλὰ as a “global marker of contrast”, one that “introduces a
correction of the expectation created by the first conjunct; an incorrect expectation is
cancelled and a proper expectation is put in its place.”" If a writer uses ἀλλα he's pointing at the contrast.
δε is more delicate than ἀλλα. It doesn't point out like "look, it's this, not that". There's no one correct translation for δε - it depends on the context. Often it can be translated as 'but', but not always. At discourse level δε is used to mark "development", for example new scene in narrative. 'But' wouldn't be a natural translation for it.
For example:
The cat is black - the dog is brown.
The cat is black and the dog is brown.
The cat is black but the dog is brown.
The information in all these clauses is the same, and there's contrast in all the clauses, whether 'but' is used or not. 'But' only makes the contrast explicit - the dog isn't black, it's brown. Whether you want to use it or not depends on the context, just like in your own example. ἀλλα sounds natural there, but only if the larger context supports its message.