Something like "distributed cache" is valid jargon. I already conceded that it can be useful. But the majority of it (by raw numbers) is the kind of stuff of the OP is lampooning -- business and office jargon. Of course there is plenty of scientific and mathematical jargon that's legitimate shorthand.

Even there, however, the line blurs. That is, you have terms with legitimate use that were poorly chosen. Sometimes the poor choice is historical accident, but often it's motivated by a desire to sound more impressive and complicated that it is. Something like "applicative functor" might fall into this category.