I'm talking about what comes before. Most don't have the luxury of a publishing department producing a clean pdf from what's no doubt a handful of Word inputs. Apart from horrendous formatting issues introduced by incompetency and variety of Word versions (including limited mobile/online versions), the purpose would be astronomically better change integration when 10 people concurrently review for starters. If you've never had to integrate changes from a handful of people for 100+ page documents, praise your deity.
The trick is to use open source version tracking and a simpler text editor. I suspect that markdown in a gui editor could do everything that actually needs doing.
Every employee with a reasonably recent office suite can produce a high quality PDF of whatever they are producing.
I actually have done such things. It’s not that bad - the process of review is the painful part.
Nobody has the skillsets to manage latex source code in a collaborative environment. I actually did a pilot a few years ago with Markdown and GitHub to collaborate on a large report for a multi-organization project. We quickly abandoned it for Word/Sharepoint and it worked fine.
The only practical exceptions I can think of are typesetting for publishing and bill drafting. Both are narrow use cases with specialized staff and tools. I definitely see where having a system output LaTeX could be useful for those use cases. But not the the collaborative aspects.
The crazy dogma around these issues hurt everyone.
The smart move is the pragmatic one - just like I can access any of the billions of US Federal Court filings in the intended presentation, mandate production of documents in a standards based PDF or PDF/A format. Then history gets accessible documents for eternity, and entities can choose to use what application makes sense for them.
Or we can bikeshed over the format dick-measuring contest.