I'll start with one fact. In the last thirty-six months there has only been one accident in the U.S. that resulted in deaths onboard -- the Delta Connection crash at Lexington, KY, Aug 6, 2006. US airlines have operated millions of flights in that period without incident. From the article above: "Crashes have dropped 65 percent,with a rate of about 1 fatality in about 4.5 million departures." Forty seven people died in the Lexington crash. How many people have died on America's highways in the past three years? Over one hundred thousand.
Ok, end of my little rant.

My point is, air travel is exceedingly safe. I think people freak out more about flying than driving because they give up control. That, coupled with this "zero tolerance" notion about so many things, and the "spectacular" and rareness factor that still surrounds an air crash. But I digress...
I feel that CNN article referenced above makes some valid points, however it takes some of the information out of context, and skews the writing to make it more sensational than it needs to be. It fails to use hard numbers where they would help. I am also strongly opposed to the hiding/destroying of survey data. But this is a political issue and I'm not much into politics. I will say that pilots, like anyone, would probably be more forthright in an anonymous survey than when reporting via more formal channels.
from the article: "the pilots reported at least twice as many bird strikes, near mid-air collisions and runway incursions as other government monitoring systems show." -- ok, but how many bird strikes is that? If there was ONE per year for the million flights that occur, that'd mean this study showed two. If it was 10,000 in the old study, then it would be 20,000, a big difference. I don't know what the actual numbers are.
I do know bird strikes are somewhat common, but the vast majority cause no damage or safety concern. In 99% of all bird strikes, passengers wouldn't even know it. Since 1975, only five commercial jet accidents have been attributed to a bird strike, and only one of those accidents resulted in a fatality. Personally, in my 1000+ hours flying as a professional pilot in the US for the past two years, I've hit two birds. Neither caused any damage or any safety concern. Of note: both were in my corporate jet at small airports. I've not yet hit a bird in my commercial airline jet. Commercial airports use air canons to make noise and disperse birds, for example.
I will say that "runway incursions" are probably the most dangerous incidents in modern aviation. The most dangerous are when one airplane crosses a runway on which another is taking off or landing. However an incident as benign as a truck crossing without authorization, even though no planes are near the runway, is still an "incursion."
There has been immense improvement in this area in the past few years with the implementation on a number of new systems: Radar tracking of planes on the ground, new lighting to better mark runway intersections, and automatic systems that sense a plane on a runway and signal a plane about to land that the runway isn't clear.
Looking at the FAA's website, it shows about 330 "runway incursions" per year for the last five years. The number has been holding steady while the amount of traffic has increase by at least 30%, if not more. So the "rate" of incursions is actually decreasing.
from the article: "The survey also revealed higher-than-expected numbers of pilots who experienced "in-close approach changes" -- potentially dangerous, last-minute instructions to alter landing plans." --
"Higher than expected" -- one percent higher? 100 percent higher? Which is it? And what is a "close in approach change?" 1 mile out? 5 miles? 30 miles? I honestly don't know. I *do* know though, that if a controller issues an instruction a pilot doesn't like, the pilot can deny the instruction. If all else fails, the pilot simply adds power, climbs away, and circles around for another approach. It costs some time and gas, but is perfectly safe- often the safest choice, and pilots aren't afraid to do so when necessary.
In closing, I'll reiterate that flying is exceedingly safe. Yes, much of the radar technology in the US is old. But there are robust policies, procedures, and standards in place to ensure safety. And most every pilot and controller is incredibly qualified and competent to do their job with safety at the forefront of their minds. It's the standards and the people that really make it so safe, technology just helps out.
The main reason for most the horribly long delays is because of these incredible safety standards that limit how close planes can be to each other, to weather, etc. The FAA limits the number of flights per hour that can depart a certain airport (JFK, for example) Airline management then goes out and schedules more flights than the FAA limit. This guarantees delays, even in perfect weather. Throw in some weather, and you know the result - hours long delays. Yes, new technology will help allow more planes to fly in the same airspace. But until that happens, we have a problem with more people wanting to fly than airspace allows. Same problem we have on congested highways in so many large cities.
I fully understand the desire to reduce the number of air crashes to zero. But what I don't understand is why this same fanaticism isn't applied to highway deaths, medical malpractice, or the myriad other ways we can die in this world.