Both efficient and potentially a problem

From Compensatory conspicuous communication: Low status increases jargon use by Zachariah C. Brown, Eric M. Anicich, and Adam D. Galinskya.  From the Abstract, emphasis added.  

Jargon is commonly used to efficiently communicate and signal group membership. We propose that jargon use also serves a status compensation function. We first define jargon and distinguish it from slang and technical language. Nine studies, including experiments and archival data analyses, test whether low status increases jargon use. Analyses of 64,000 dissertations found that titles produced by authors from lower-status schools included more jargon than titles from higher-status school authors. Experimental manipulations established that low status causally increases jargon use, even in live conversations. Statistical mediation and experimental-causal-chain analyses demonstrated that the low status → jargon effect is driven by increased concern with audience evaluations over conversational clarity. Additional archival and experimental evidence found that acronyms and legalese serve a similar status-compensation function as other forms of jargon (e.g., complex language). These findings establish a new driver of jargon use and demonstrate that communication, like consumption, can be both compensatory and conspicuous. 

I am of mixed mind here.  I simultaneously believe that jargon is frequently used for obfuscation purposes to bamboozle others.  I am comfortable accepting the proposition that there is a status signaling element as well (though probably needing some more robust replications than this paper alone.)  I am also comfortable with the proposition that jargon can be used as an exclusionary device.

On the other hand, jargon can be enormously useful.  Irritated with the amount of jargon in a report we were producing many years ago, I self-righteously undertook to re-write it in plain English.  With mixed results.

The target audience of the report were informed and knowledgeable individuals familiar with the topic.  I was able to simplify some of the jargon, but not nearly as much as I anticipated.  Jargon is often simply a shortcut for those with a shared body of knowledge.  Some jargon you can replace, but usually at the expense of making the report dramatically longer.

In these situations, the jargon is less an explicit word than it is a call on a concept.  Invoking an S-curve (or sigmoid curve) or a Pareto distribution or the Peter Principle or any other shared idea really functions as a call on a set of abstract concepts.  It is a signal that "this is the frame within which I am speaking."  As such, it is an allusion that directs the readers attention to a concept rather than a direct explication.  And as such, it can be very efficient and succinct.

Like so many things, jargon is both efficient and potentially a problem.  It depends on context as to which it is and to what degree.