If you're just typing simple text documents, the current Word is not going to save a massive amount of time over WordStar, but it will save some thanks to spelling corrections etc. And even a few seconds per hour per worker can represent a lot of value for large organizations.
If you're doing an essay or a brochure with headlines in different fonts, double-column text, embedded pictures, math equations etc, then the time savings will be massive. Just the ability to move a pic while text flows around it provides huge benefits.
When my son was in his early teens, I watched him use Microsoft Office to create complex documents that would have taken an IBM PROFS mainframe user weeks, or might have been impossible. His homework looked more professional than IBM's corporate documents.
I guess there are plenty of workers who use Word at the same level as they used WordStar, or who reckon that (back on topic!) LO is just as good. (Yes, LO is almost as good as Word if you just want to replace WordStar.)
But are those our criteria? Should we ignore the fact that people who have open minds and a willingness to explore (like children) can quickly and easily do new things that they can't?
I'd prefer to have a product with a crappy UI that can interoperate with everybody else seamlessly, than one with a great, simple UI that can't.
There are groups that investigate the risks. EUSPRIG (European Spreadsheet Risk interest Group) has some great documents. http://www.eusprig.org/quotes.htm
I don't know if anyone is testing the main spreadsheet software with fuzzer-like tools; or if there are any mass audits of spreedsheets. Both would be useful.
So, it doesn't work now?
The main question would be 'how much more productive one can be now than before?'. It would be nice if there were independent quantative studies on this - as word-processor use is a central part of most governments administrative functions I'd imagine that one can't move for the vast number of such studies ... lol.
I'm guessing that the UI changes between Wordstar and Office 2015 amount to a tiny change in the time needed both for training and useful production of word-processed documents. When it comes to spreadsheets I can see things have probably moved on - at least in speed of processing if nothing else.
I'd love to ask, of the top 100 functions used - how many weren't in office applications 5 years ago, 10 years ago? How has the average time per word in production of business letters changed (reduced?) over time and what specific [versions of] applications caused the greatest change?