I think you’re glossing over the ultimately philosophical argument here.
There is a fundamental argument between descriptivism and prescriptivism. English teaching is traditionally prescriptivist - that is, it involves declaring certain usages correct and others incorrect, regardless of how common the ‘incorrect’ usages may be in the wild. At its most ridiculous extreme, in the seventeenth and eighteenth centuries misguided principles were offered up on the basis of comparisons to Latin or arguments from logic. Thus we have the traditional rule against “splitting infinitives”, based on a particularly bad analogy to Latin grammar, and the injunction against ending sentences with prepositions, because doing so is impossible in Latin. Even though those rules aren’t taught much anymore, they’re still mostly avoided in formal writing.
Descriptivism, on the other hand, is the attitude underlying linguistics, since linguistics (ideally, at least) is the scientific study of language. Obviously, declaring certain usages ‘correct’ and others ‘incorrect’ is incompatible with scientific study; the notion is rather absurd on its face from the perspective of a linguist. It would be like Jane Goodall trying to teach the chimpanzees the value of chastity - science is inherently based around observation, not legislation. I’d go into exactly why I think prescriptivism is inherently problematic, beyond not being useful in academic study, but this post is way too long as it is.
Dictionaries used to prescribe certain uses (in fact, many of them still contain usage notes - but their recommendations are generally not in line with traditional prescriptivist commandments.) Webster’s first dictionary, for instance, deliberately promulgated reformed spellings, and it’s to Noah Webster that we owe the American spellings of color and center. Other spellings he recommended never caught on, though - you won’t find many formal writing with thru or tho, tho both are occasionally found in informal writing. Modern dictionaries, in contrast to those of Webster’s day, are basically descriptivist in their approach, functioning as documentation of usage and normally including nonstandard words and pronuciations and meanings that are disliked by prescriptivists. Note that, considering the way dictionaries are designed, it’s more accurate to refer to them as documentation of language, rather than authorities to be consulted. Note the use of scare quotes in the OED quotation you include; the OED documented both the common use and the particular uses recommended by prescriptivists, but they weren’t particularly recommending any usage - just describing the various ones in existence.
Neither Case Sensitive nor I are particularly inclined to accept the dictionary’s word when it comes to usage advice, I suspect. The OED is a work of tremendous scholarship - it’s the most complete and most thoroughly researched work of lexicography in existence. But it doesn’t purport to represent itself as a usage guide, and it shouldn’t. That’s not within the realm of academic study of language. And those who believe in the silly flights of prescriptivist fancy that Case Sensitive has engaged in here - notably, the idea that if people handle it roughly, language is liable to break on us - are not usually friendly to the idea of dictionaries as documents rather than sets of rules.
You seem to be approaching this from the common, but fundamentally mistaken, perspective that there is some ultimately ‘correct’ form of language, some ‘Platonic form’ of the English language that truly exists somewhere out there, but that can never be realized in our imperfect universe. This perspective is mostly promulgated, in my experience, by schoolteachers who simply describe some uses as “bad English” or “improper grammar”, as though a child’s speech has been objectively measured against the Platonic ideal and found wanting. But there is no abstractly ‘correct’ language - there’s only the language that people actually speak, in the various circumstances they speak it, and the particular, arbitrary, rules that the prescriptivists yell about. In many cases, we can even trace the origins of those rules, and figure out exactly who came up with them in the first place. When people talk about a particular sentence being “technically incorrect”, they’re commiting this particular error. There is no such thing as “technical correctness”; you can attempt to guess at whether or not a sentence is the sort of thing that people actually say (and that’s what linguists do, with the imperfect tools we have to do it), or you can say that it passes muster with some particular prescriptivist authority, but - given that the prescriptivists themselves squabble about what is “right” and what is “wrong” - it’s simply logically invalid to imply that correctness can be measured “technically”, rather than simply being a matter of opinion.
A lot of people regard dictionaries as authorities on correct usages, but the authors of dictionaries don’t represent them as such, and it’s a mistake to imbue them with any particular authority in the matter. It simply doesn’t make sense to try to guess at gradations of “correctness” and “incorrectness” as reported in dictionaries, because it presupposes some objectively correct or incorrect usage that dictionaries attempt to convey. And that Platonic form of the language, as I’ve pointed out, simply doesn’t exist.