I started using 
during my undergraduate studies (late 90s). In those days, MS Word had a bad tendency of dying suddenly and taking with it all of the work you had done in the last hours. Also, using the equation editor those days was banned by the Geneva convention because it was worse than torture. And the final results produced by where always way nicer than what you could get from Word, no matter how hard you tried.
Years passed and Word improved a lot. Not only it didn’t crash, but making documents look like you wanted them to look became easier every time. I went back to Word for my MSc thesis for these reasons (and also because my advisor loves to review documents using Word). This worked out pretty well, even so that I also started using Word to write a journal papers (one which got published).
But here I am going back to . The reason is simple: a paper I sent to a conference was rejected, and now I want to improve it and send it to another conference. But alas! each conference has a difference document template! So I basically have to re-write my previous paper in the new template… But in this is not the case – I can take the text, change the formatting, change the styles and voila! (there is more to it than just this, but it is a LOT simpler than trying to copy and paste between two Word documents). This is where the philosophy that the content should be separated from the typesetting saves you a lot of work. BTW, this is similar to the “Separation of Concerns” principle that is so much valued in software development. So beware , here I come!
Related
Welcome back to LaTeX! IMHO, it is the only adequate tool for authoring scientific work.
However, having worked in scientific publishing for a good while, I know that submission processes – with the constant rise of electronic submission systems – are still hindering the consistent use of LaTeX: a lot of scientific journals still request authors to submit Word documents. This is due to the fact that there still aren’t automatic converters good enough to reliably cope with the potential heterogeneity of .tex documents. This is where – full circle – software developers (especially web developers) are needed.
Of course this issue – amongst other factors – is inherent to the open, ever-evolving nature of an open source product/community such as LaTeX(‘s). However, I think this is something scientific publishing really needs to tackle, if LaTeX is at some point to take on the role I think it should have: that of a standard tool for scientists. Until then, however, we’ll just have to submit compiled PDFs along with .tex files. Or resort to the use of Word…