Nielsen Media Research has long been that sleepy and mysterious enterprise — occasionally vilified, frequently cajoled and periodically challenged by would-be competitors — whose findings determine whether TV shows live or die.

Lately, however, the company has roused itself to pursue grander ambitions, just at the point where exploding media options and the hunger for easily digestible data suggest the TV industry might benefit if Nielsen stuck to its knitting.

Acquired in 2001 by Dutch giant VNU, Nielsen changed hands again when a clutch of private equity firms took control last year. Subsequently renamed the Nielsen Co. (whose assets include such publications as the Hollywood Reporter and Billboard), the outfit has embarked on a series of initiatives punctuated by what seems like a new-venture press release of the week.

Beyond augmenting its existing measurement system by, among other things, agreeing to rate commercials, Nielsen’s peripheral endeavors — documented in a slew of recent announcements — include establishing a “global ratings metric” for in-store marketing and testing something called Hey! Nielsen, billed as a “social network site for a diverse range of pop culture enthusiasts who are eager to make their voices heard.”

Apparently, Geek!-Speak didn’t clear legal.

The flagship business still hinges on the TV ratings service that operates out of Florida, where number-crunching gnomes sift through 10 million viewing minutes a day. And while Nielsen is clearly looking to enhance its profile by addressing brave new worlds — helping make sense of downloads and uploads, broadband and narrowcasting — at times, it seems the old one is becoming the wild West of spin and obfuscation.

Even traditional ratings generally remain a source of bewilderment outside of TV research departments, where executives and producers struggle to remember the distinction between a rating and a share. The ad-driven emphasis on demographic data — expanded to encompass pie-slicing categories such as “boomers” (as in baby boomers) or “tweens” (9-to-14 year olds) — only exacerbates the confusion.

The proliferation of information has also emboldened companies to play fast and loose with Nielsen data, leading to a spate of questionable or misleading proclamations — including three major groaners in September alone.

Univision began the “oh no you didn’t” derby by reporting a cumulative audience for its dubbed-into-Spanish presidential debate, claiming that the net’s 4.6 million viewers surpassed previous debates on ABC and the cable news networks.

But whoops, that number (commonly known as a “cume”) counts anybody that watched the debate for even a few minutes, as opposed to the customary measure of the average audience at any given moment, which slashed Univision’s total by 54% and dropped its debate to sixth among those televised.

PBS — an unlikely culprit to engage in such sleight of hand — was the next to transgress, not only using the cume to state that 18.7 million people watched “The War,” Ken Burns’ handsome World War II documentary, but further inflating its results by combining two airings of the premiere. Held against a more conventional yardstick, the audience actually plummeted to a respectable but far less imposing 7.3 million.

Finally, an arcane new Nielsen rule frees networks to include repeat airings in a program’s reported audience, providing the ads are the same. NBC quickly availed itself of this opportunity — to howls of derision from rivals — with this season’s premiere of “Heroes,” absorbing a low-rated Saturday encore into the firstrun Monday telecast, thus basically eliminating an average-deflating hour from its broadcast week.

Nielsen has never enjoyed mediating intramural disputes or slapping clients around for distorting its data. Still, inasmuch as the ratings are not just the industry’s currency but serve as a cultural barometer of popular taste, everybody suffers when bogus inferences and dubious spin float into the ether — making the numbers look more untrustworthy, and feeding the suspicions of conspiracy-minded yahoos already convinced that black-suited government agents are seeking to deprive them of “Jericho” or “John From Cincinnati.”

The irony is that Nielsen deserves kudos for a new plan to expand its sample size — tripling the pool to 37,000 homes and 100,000 people by 2011 — recognizing the need for more info to adequately reflect a fragmented audience. As it stands, if a few loyal Nielsen-monitored Food Network or Golf Channel viewers leave town the same week, the network’s ratings can plummet.

Nobody ever loves the referee, which is ultimately Nielsen’s thankless role — functioning as the impartial arbiter in an incredibly high-stakes game. So while the company understandably has its own business imperatives and objectives, they won’t amount to much in the long run if pushing into new arenas takes Nielsen’s eye off the ball.