Language is a means of communication, and as such the whole idea of it is of course to get the message across, to make youself understood. No one is suggesting that we start talking gibberish to each other, the question is rather who decides how we talk, what grammer is, and how it changes.
You have a point here. However, I think it is well known that no texts are as hopelessly impossible to understand as sceintific texts and law! Scientists are often very bad at describing what they mean, and as for law texts, you need lawers to translate them and they argue endlessly over them.The necessity for a correct usage is most important where precision of meaning is required: scientific papers, enactments of laws, interpretation of contracts and so on. If it were not possible to describe a scientific experiment in precise terms, then it would be difficult to repeat it; if a law were expressed in vague terms, it could not be enforced; if a contract were unclear about what had been agreed to, how would you know if it had been properly performed or not?
Agreed.This is not to say language must be codified and set in stone. Only a dead language does not change - cf. Latin. In any living language, attempts to prevent change, or to prevent certain types of change, will prove futile and are doomed to fail, even with government support, as with L'Académie française, which is a self-aggrandising body, overly conservative in outlook, and ineffective in what it does.