In biology (in government and industry) we use both. And we tend to use Word because that's what everyone else is using. But WordPerfect and LibreOffice generate TOC and figure lists, and track changes just fine. LaTeX and git also accomplish those goals. But the problem is working with other people in different groups, companies, levels of expertise, preference, etc. When track changes is breaking between LibreOffice and Word you just revert to using word to find the changes, and then incorporate them by hand into your LibreOffice document. Or, since you're using Word anyway and it does accomplish your goals, you just switch to Word.

Working in any setting aside from a self-contained research unit or the academic world, you just learn Word. In my case I learned Word and I use all of the various variable-esque and markup I can to make generating TOC, figure/table lists and references and adjusting values easier when working with people who just use Word because that's what you write reports in.

Even in an academic setting, if your advisers use Word (and prefer Word over dead trees) you're going to be creating Word documents for your thesis and academic papers (the journals mostly don't care because they re-typeset everything anyway) because you need their expertise to make your work awesome.

So, what about LaTeX? I like the idea more than the practice. Making basic changes (or what would be basic with CSS, Word, InDesign) to templates is mindbogglingly complicated to beginners. I'm not sure it's even particularly easy to create custom templates and formatting for experienced LaTeX users. However, once you have the Ecology, Washington Department of Fish And Wildlife, USDA, etc. LaTeX template all future publications for those is much faster.

If you want to know the actual one true solution to kill Word for academic, industry and government publishing it's this:

Modify Lyx (or make something like it using Markdown or reST) that uses sane defaults (I think WYSIWYM with a live preview is probably okay for everyone except the "lost dog" folks), some auto-correction, spelling and grammar checking tools and has the ability to easily and correctly make headings, adjust page headers and footers, footnotes, endnotes, and insert and edit figures, tables (the table formatting tools should be very powerful, ideally also the ability to include data from Excel and also tie into R, SAS and SPSS), images and variables (for things like n, p-value, species names) and integrate it with Evernote, Endnote, Papers, Mendeley and BibTeX (though, supporting BibTeX gets you pretty far). Now, make an extensive repository of formats for publications (like PyPI, rubygems, etc. but for templates from academic journals etc.) Ideally the templating should be powerful enough for experts to create anything they can imagine and simple enough that the Masters student studying the behavior of Little Auks can get her short fact sheet to look how she wants it to for her presentation to the Reykjavik Ornithological Society (using system fonts should not be difficult or require any amount of documentation reading).

Now you need to add collaboration tools. Track changes, highlighting, and commenting gets you to 90% of use cases for Word. Add in version control that is easy and makes sense, the ability for multiple people to live edit the same document and good merging tools (for multiple offline changes to the same document by different people) and suddenly your software is more useful than Word.

Oh, and it needs to be open-source, because us science-ey folks are increasingly grumpy about closed-source things, and we already have to buy the Microsoft Office suite to deal with documents and data from multiple different sources. It also can't be multiple gigabytes to download because I need to be able to tell my colleagues, "Hey, you should use this, it's easier to make pretty things, and the collaboration tools are more useful than Word's."

But even that only gets you the people who are writing for (essentially) print. If you have output formats that create sane, semantically correct and useful HTML (for people writing for online magazines, blogs and static pages) and HTML/CSS for people who want to use it as web publishing system you've got a tool that could slowly start gaining a foothold.

"A lot of software developers are seduced by the old '80/20' rule. It seems to make a lot of sense: 80% of the people use 20% of the features. So you convince yourself that you only need to implement 20% of the features, and you can still sell 80% as many copies.

"Unfortunately, it's never the same 20%. Everybody uses a different set of features. In the last 10 years I have probably heard of dozens of companies who, determined not to learn from each other, tried to release 'lite' word processors that only implement 20% of the features. This story is as old as the PC."

Does she NEED Word? Technically not I suppose, but Word is the tool that makes her job the easiest. That being said she still curses Word and Endnote all the time, so there is no doubt room for improvement.

In the Engineering/Consulting industry we use it often to collaborate on proposals, reports and other documents. Key clients - many of them utilities or energy companies - also prefer Word documents and use Track Changes.

It's very hard to move away from that momentum, partly because you can't use another tool and just export to .doc/.docx due to the format being so arcane.

For what it's worth though, Track Changes is what Word does pretty well. I rarely use it outside of work though, preferring something like Byword on my Mac or just TextEdit.

Why do you need any more? Legal itself is a huge field and a lot of the work produced are obviously documents. Even if it's just the legal field that use Track Changes, it seems like it would be a huge win.

It's not like Tracking is forced upon anyone. It's an extra feature that stays out of the way unless you want it there. I dont understand why this is even an issue.