Nope, because it doesn't work. Now acrobat (this is windows) has gotten better in this respect over the years but if your PDF had any sort of kerning done with the text Acrobat would get the word breaks wrong. The cut-text algorithm was based on the assumption that words were dropped in, well, word sized chunks. (I say was since I have not used acrobat for awhile and things might have changed.) A paragraph typeset by something like TeX would typeset things a letter at a time to do kerning and also to slightly adjust the kerning to optimize the white space when justifying a given line. It typesets beautifully, but can really throw off an algorithm making the assumption that the pdf came from Microsoft Word.
Also if you think back, it wasn't many years ago that you could not make any electronic submission. Then when they started taking them they had to deal with Word, Word Perfect, RTF, Postscript and whatnot (I distinctly remember that PDF was not an option for the first electronic resume I submitted, I used Postscript). The secretary printing and OCRing them approach probably made a lot of sense at the time. Also you cannot assume that the application was something that could be copied into, I would not be surprised if it was basically a black box.
The reasons for doing lots of things have been lost in the not so distant past. It is not always the case that the people doing something which looks like it has an easier approach today are sticking with their old stuff because they are just stupid. It could just be that it is cheeper to keep paying for the labor and toner than to buy a new system and retrain all the other employes.
Legacy systems are funny that way.