Technically, it might be more accurate to say that HF pirate radio activity can only be roughly estimated due to the inherently imprecise nature of the available data.
Some broadcasts are heard but never publicly logged at all. Over the past few years of lurking nearly 24/7 on IRC, I see many mentions of shortwave pirate broadcasts heard by regulars on IRC, but which aren't logged to the HFU, FRN, FRC, FRW or Alfa Lima. Since early 2011 I haven't posted public logs nearly as often as I used to, although I'll usually mention the activity on IRC just to give others a heads up. I'd guesstimate I log fewer than half of the broadcasts I actually hear now, mostly due to time constraints.
And I've read comments from various regular DXers in the shortwave pirate scene who may e-mail a station operator with loggings, but who may not always follow up with a public post to the major pirate radio hobbyist forums from which the data was harvested.
Another complicating factor: apparently some HF pirate loggings from the period covered were of dubious authenticity, if accusations of sock puppetry, self-logging and even faked loggings (of non-existent broadcasts) are credible. It would be difficult to estimate whether such dubious loggings were statistically significant enough to skew the available data.
Yet another complicating factor, although probably insignificant in terms of overall data: some loggings to the FRN were deleted in 2011, and others were omitted from the FRW. This occurred only during a relatively brief period of time during early 2011 and probably accounted for fewer than a dozen deleted and/or omitted loggings, mostly of Northern Relay Service.
Even allowing for the fudge factor in loggings, the available data is unlikely to reflect actual activity because of erratic propagation the past couple of years, which seems to primarily favor regional reception within a few hundred miles at best. There's been plenty of activity west and north of the traditional Ohio Valley territory, but it's not always reflected commensurately in loggings because there seems to be fewer listeners posting public logs.
My general impression is that there was somewhat less North American shortwave pirate activity in 2011 compared with 2008-2010, and also a smaller percentage of those broadcasts logged. But this is only a rough guesstimate based on eyeballing chat logs daily and checking publicly posted logs almost daily.
From the referenced article: "Enforcement against AM and shortwave broadcasters is also on the rise - this is most likely due to an increase in activity on these bands rather than a change in FCC priorities regarding unlicensed broadcasting."
This speculation about increased activity is not supported by the facts.
Ragnar's list of monthly activity for January through May, 2011 (https://spreadsheets.google.com/oimg?key=0Asx79vbD_EnqdHlVQXBQRWpHOEVLVXd1TGJEaERNTkE&oid=10&zx=qai8il-5a5fvl) shows approximately 260 broadcasts (data is not available for the rest of 2011).
During the same period in 2010, there were approxiamtely 385 broadcasts https://spreadsheets.google.com/pub?key=0AiCFvPP_I4cOdEFwQ1lrOFlpUnJMQ3VqUFJPYkdVd3c&hl=en&single=true&gid=9&output=html.
This was actually a decline of 33% in activity.