I started off studying four-part harmony the way it is generally presented. I even tried to make sure I'd learn it deeply enough by asking for a book that was not just geared toward passing exams. Eventually I ended up with four or five texts, because each left out something crucial. A year or so into, and having stopped lessons, I met my present teacher, who warned me that this material would put music into a box. Nonetheless we went through it. One of my main books was Horwood, which was more demanding and had you create 8 measure lines to begin with.
The first thing we found was that the rules prevented you from many things that would make good sounding music. It had to be restrictive, since it was aimed at giving students rules when their experience was still limited. Actually listening to how your music sounded was barely on the radar, so if you were a good drone, and followed the rules, you would produce something that would pass the muster of exams. But if you were a musician, you could not use your ear or instinct, because of the rules.
The next discovery was that Bach himself did not follow those rules, even though they tend to be ascribed to him. Later in one of the official books used these days, I found small print for teachers, stating that everything had been simplified, and that any teacher should feel free to teach more broadly according to their judgment.
One book that was used by the RCM but has been discontinued after the last revision of the syllabus, tried to get past the narrowness. It stated at the beginning that the rules applied to a restricted period of music history, but not before or after - i.e. especially during the Baroque period (notwithstanding that Bach himself went beyond those rules). It began delightfully with some general principles that transcend those narrow rules:
1. the tritone, of which there are only 6 if you discount inversions and alternate names
2. the "up P5, down P4" pattern, and what this implies.
Thereafter they invited exploration of these two things, especially the tritone, regardless of what official rules they had to teach, in order to keep the door open.
The book also tried to promote musicality - which, however, could not always be done - because of the narrow restrictions of the rules. The exercises themselves, written to stay within the rules, were often too dry to allow for musicality.
This last book also tried to give examples from actual music, and here at times it was "tire par les cheveux" - something which was actually in the middle of a transition, with a tiny section pulled out.
There is the "dominant 7" which, if you take it as being a major triad topped by a minor 7th in root position, does not always function as a dominant. But in these studies it always does. In some music, for example, you can have a series of these chords, which do not resolve in that way and do not have that function. In other times, if you go by ear, you will have your "augmented 6th" chords, which the theory does cover but makes mysterious and full of rules.
When you get to later music, it gets really complicated if you try to look at music only according to this system.
The final conclusion was that "4 part harmony" seemed to be a specialized study of a narrow period of time which for some reason is taught right at the beginning. I am still ambivalent about it. I sort of enjoyed having my writing sound like Bach's amateurishly weak little sister.
Topic: Why is Four-Part Harmony taught in Music Schools?