Can you provide evidence of this? Given the number of places I've found pretty much exactly the same thing being stated, I'd love to see evidence to the contrary.
Any study of Linguistics will tell you that any known and repeated structure in a language is the observation of custom not of rules. There is no Congress of Languages that passes a Law of Tongues, no first principle of language that requires things to be conjugated this way or ordered in this way. Language is a temporary contract between the speaker and the listener. A protocol that if I use these sounds in this way, I have a pretty good notion that information will be transferred to the listener.
Over time, as needs change, the contract changes and what I need to say to transfer information changes. More importantly, I have to keep track of several contracts because not all listeners have signed the same one! This is called code-switching, and it's why I talk to my boss differently than to my much younger niece than to my mechanic.
Language prescriptivists don't seem to understand that and will refer to their favorite dictionary and grammar guide (usually published sometime between the 18th and 19th centuries and written by people of the same ilk trying desperately to coax order from chaos) as if they contained the great laws handed down from the God of Languages and will try and convert the World by force to sign the contract that they think all the world should be on.
None of which suggests that the stated rules (or "guidelines", if you prefer) are "more opinion than fact". Yes, they are not as strict as rules of, say, mathematics; and you won't be thrown in jail by the grammar police for braking them. However, there is a customary way of ordering adjectives, and ignoring it will make you sound less natural.
I am in a strange (though these days perhaps not that unusual) situation of speaking English as a second language but being more comfortable with it than with my native tongue. I was never in a primary school and was never officially taught any grammar rules. Yet to me, "red big coat" sounds as strange as it would to a native speaker.
A language is defined by its usage, and the rules/guidelines do shift with time. However, at any particular point there are definitely a "right" and a "wrong" way to construct a sentence. Therefore these rules (weak as they are) are definitely not just "opinion" that can be freely ignored. A rule defined by custom is still a rule.
Nobody said you have to or even should ignore them, just understand that they aren't hard and fast rules. It's incumbent on the language producer to understand the customs that need to be observed to make themselves understood. Not following the custom will seem strange at best and at worst will prevent you from being understood. That's all there is to it.
What is opinion is that some people think that there are hard rules to language and it's simply not the case...it's their opinion that there are such rules, even if fact doesn't support their opinion.
You seem to have a strange definition of "opinion" ...
> What is opinion is that some people think
> that there are hard rules to language ...
No, it is a fact that some people think that there are hard rules to language.
But all this is beside the point. Yes, you are absolutely right that language is fluid, changing, different in different places, and trying to produce "proper rules" and "proper grammars" (by some hard-to-define meaning of the word "proper") is like nailing jello to a wall. You can try, but you won't have much luck.
That being said, there are some "rules" which, if you follow them, help you to communicate effectively most of the time. There will be exceptions, and there will be geographical variances, and these are not rules of the language, but rather, they are rules for you to follow while synthesizing constructions.
To be honest, I expect you are in violent agreement, and arguing with attitudes you have found elsewhere, but which I do not hold.
It is a fact that some people have the opinion that there are hard rules to language. Better?
More seriously, I think we're in violent agreement, the difficulty is how we parse the word "rule".
If rule means "something that must be done or you'll be a bad person and the world will end" as many language prescriptivists like to use, then it is a fact that languages do not posses rules of this sort.
If rule means "a mutual set of understood guidelines" then yes, I'm in full agreement that languages can have sets of customary rules that aid construction and understanding, but this set of guidelines is incredibly fluid and can change as you've described including temporal, generational, socioeconomic, ethnic, fluency and other sources of variation that can alter the set so profoundly that two speakers of the same language can be more or less unintelligible to each other yet still be said to be speaking the same language!
Back on topic: "a green great dragon" sounds awkward unless we're having a discussion about the colors of "great dragons" in which case it's perfectly ordered and "a great green dragon" makes no sense since we're not talking about sizes of "green dragons".
However, the original question is interesting even from a descriptive point of view. Essentially, the descriptive version of the question would ask what is a simple set of rules that accurately model the ordering of adjectives in English (or perhaps in a specific dialect/register of English).
Thanks.