In the last few NFL seasons, we have not had what one could possibly call "parity." Particularly in 2005, as sports writers such as Dr. Z and Bill Simmons pointed out, you had a big chunk of good teams, a big chunk of bad teams, and not a lot of mediocre teams. This isn't exactly Pete Rozelle's dream of "any given Sunday" with everybody finishing around 8-8. 2005 featured the opposite of parity--you had a lot of good teams who, most weeks, faced far inferior competition and easily beat them. 2004 featured much the same thing. Instead of a bell curve, you have a dumbell curve.
And at the top, as I attempted to show below, we don't see the signs of parity. Teams that make it to the Super Bowl usually do so during the beginning, middle, or end of a successful run of contention. That is what is important, I think. It's good that the current system allows teams to rebuild quickly; the bad teams can get good again quickly. But I don't see any evidence that the Champions are diluted in today's NFL. Indeed, the 2003 Patriots went 10-0 against teams with 10+ wins. I think that team (and the '04 team) could compete with Lombardi's Packers, Noll's Steelers, or Walsh's 49ers. I don't know that they were better than those teams...but they would compete.
That said, there are some signs of parity. Looking through records at pro-football-reference.com, I struggled with the methodology I should use. I didn't want to separate eras arbitrarily to compare, because the cutoff would alter the results (for example, if I want to look at the number of teams that make the playoffs in a given era, and I decide to cut off eras at '70-'79 and '80-89, I can't deal accurately with a team that might have made the playoffs in '79 and '80 because they would each count for one playoff appearance in each era, when in actuality those appearances should be looked at together). Furthermore, since the league merger in 1970, expansion of the league and the playoff format makes using playoff appearances difficult. From '76-'77, only 29% of teams in the league made the playoffs. From '90-'94, 43% of teams in the league made the playoffs (Peter King provides a nice graph). Today, 38% of teams make the playoffs, compared to 31% at the merger.
I decided to examine a certain phenomenon in assessing the "parity" of the NFL: Isolated Playoff Appearances (IPA). An IPA occurs when a team makes the playoffs, but had not made the playoffs in the previous 2 years and does not make the playoffs in the next 2 years. I examined occurances since the merger, but can only count from 1972-2003 ('72 because it doesn't make sense to try stretch the statistics back before the merger, '03 because we can't know whether '04 or '05 playoff teams can count). I excluded the 1982 season because 16 of 28 (57%) of teams made the playoffs (even though would have only added two Isolated Playoff Appearances). I included the second-year Panthers' '96 playoff appearance, because since they didn't even exist in two years previously, it's a good example of an Isolated Playoff Appearance).
I think this is a fair measuring stick for parity. We know that teams can quickly rebuild--but truly rebuilt teams won't have an Isolated Playoff Appearance, as they'll compete for several years. But in a parity-driven league, a poor team can have one decent (or lucky) year before playing like a poor team again.
My findings show that there is a certain level of parity.
1972-1979: 2 Isolated Playoff Appearances
1980-1989: 5 Isolated Playoff Appearances
1990-1999: 13 Isolated Playoff Appearances
2000-2003: 5 Isolated Playoff Appearances
In the '70s and '80s, IPAs were uncommon. The '90s saw a boom in IPAs, and the '00s seem to be continuing that trend.
There are signs of parity, but not at the level most people talk about parity.
(note: I only see usefulness in comparing the different NFL eras--it is simply not useful to compare different sport leagues. The nature of the NFL game itself (more players, more injuries, shorter careers) means that there will be a higher level of parity than, say, basketball.)