This is just tongue in cheek, of course.
I wrote my thesis in 1993, and in that time period, my anecdotal observation was that the students who used LaTeX generally took longer to finish.
Today, I use MathJax inside of markdown, and don't even use word processing very much. Industry has gotten less formal. Most text is edited in an e-mail editor with screen shots pasted in. The main use of word processing is documents that nobody will ever read, such as functional procedures and HR proclamations.
Even before this trend, there was a lot of documentation that was never intended to be read. Huge piles of what I call "ornamental" documentation, e.g., in fulfillment of ISO-9000 or other regulatory obligations.
How much of that is down to the choice of tool, and how much is down to people choosing LaTeX being more likely to be more exacting about their work overall? I think it's difficult to a say.
(FWIW, I started my thesis in LaTeX, then gave up and used Word because it was faster - but only because I didn't know LaTeX prior)
Most of the other students had whatever old MS-DOS boxes they could afford, commercial quality typesetting was impossible, so nobody demanded it. My advisor didn't even send typeset copy to journals, because he figured they would just retype it anyway. Acceptance of manuscripts in computer readable format was just on the horizon.
There was a strong tendency to get emotionally attached to the physical document -- it was the pinnacle of your career after all -- and to get wrapped up in the aesthetics. I avoided that, partially because the clock was ticking. My boss had accepted a job at another university, and I had to clear out by a certain date. Nothing sharpens your mind like a semi truck coming to haul your experiment away.
Like him, I cry myself to sleep at night for not being awesome enough to use TeX, but instead using the layman TeX, Latex.