In a post below on the history of heavy metal, TGGP uses an unfortunately standard nickname for the Maiden and Priest sound -- New Wave of British Heavy Metal, always referred to as NWOBHM. If you read any social science article, textbook, or even popularization, you've no doubt come across all manner of pointless and confusing acronyms. They must require this in social science graduate programs, just like grant writing, as a career-building skill. You can easily detect the nerdy origins of the Game community by their extensive use of acronyms (PUA, LJBF, etc.).
Yet as far as I can tell from my admittedly limited reading of pre-modern authors, they would have found this practice repellent -- both the creators and the audience. Sure, they existed, but they were no more than a handful -- AD, RIP, INRI, etc. St. Augustine did not refer to the City of God over and over as CD. Compare with the terms in modern religion, i.e. social science, again by browsing their article abstracts. Dante did not uglify his descriptor Dolce Stil Novo by calling it DSN every time after the first use. Compare with names that current critics give to today's art movements: NWOBHM, EBM, etc. And Hobbes did not market his concept of the war of all against all with the Latin acronym BOCO, perhaps because it was too pronounceable -- only real estate acronyms are designed to sound mellifluous. Compare that to one of its modern counterparts, MAD (mutually assured destruction).
Even natural scientists did not used to be geeks, unlike today. Browsing through the Wikipedia article on string theory, I see references to TOE, QCD strings, and AdS/CFT correspondence (bonus points for the slash, though it could use hyphen too, or would that be gilding the lily?). Then I look through the article on classical mechanics and find not a single one! No reference to the law of conservation of angular momentum as the LoCoAM, for instance. Newton was above using his insights into the laws of motion as a pathetic excuse to manufacture another fleet of operose acronyms.
The practice destroys clarity, introduces unnecessary jargon, and further divorces the concept from the reality it's supposed to represent -- to further Platonify it. Somehow an unspeakable array of characters seems more pure and ideal than words or phrases that roll off the tongue, like the ancient Hebrew tetragrammaton YHWH used to refer to their god. This showcases the backwardness of modern thinking, where we treat mankind's slapdash mental constructs as sacred, and the real world out there as profane. Also, they don't function merely as shibboleths, as every group makes tons of these without cranking out acronyms.
Clearly there's none of this funny business from the birth of the printed word up through the 17th C. Based on when autistic and Aspbergery traits start exploding, as seen for example by the rise of thick-skulled grammar Nazis and other top-down attempts to make human language mo, I'd guess that this mess got going during the 18th C. -- why follow the tradition of using words and phrases when it's more rational, efficient, and progressive to use acronyms? It must've really taken off during the 19th C., and of course by the end of the 20th it was deeply entrenched.
For all the rationalizations about how time-saving or ink-saving the acronym is, yet modern books blather on for much longer than pre-modern books do. Look at how easily you could condense a lot of novels compared to epic poems, and that's leaving aside that most pre-modern writings were not even as long as an epic poem -- or anywhere close. The same applies to today's non-fiction "idea books" compared to pre-modern ones. So the efficiency excuse is bogus. Semi-autistic geeks coin acronyms just to give their simple minds something easy to bat around, whether playing alone or with friends.