[T]he periods where a majority of Americans felt the country was on the right track are few and far between. The only sustained period, in fact, was during the late 90s, presumably because of the economic boom. This chart doesn't go back this far, but I'm willing to bet the rest of the 1970s were just as bad, and the late 60s not much better, which suggests that Americans haven't been happy with the direction of the country since the postwar boom, and that economic conditions had a lot--if not everything--to do with it.
I think this is probably wrong. More likely, it's an instance of the general phenomenon in public-opinion polling that when you ask people to evaluate large, nebulous, messy institutions ("Congress", "government", etc.) you tend to get very pessimistic responses. But when you ask people to evaluate more specific things (their own congressman, Medicare, etc.) you get a far rosier picture. This helps explain why, though Americans seem to always think the country is on the wrong track, most people tend to be relatively sanguine about their own personal financial situation. Of course, this doesn't mean that the right track/wrong track numbers are meaningless--the fact that the wrong-track figure is at a near-record high does suggest the obvious, namely that people are exceptionally dissatisfied with the Bush administration. But just because 82 percent of people say the country is on the wrong track doesn't imply that 82 percent of Americans think their lives suck.
There's also the interesting question of exactly why the wrong-track number is at 82 percent. It just seems hard to make the case that things in America are objectively as bad as they've been in the history of polling. The economy is sputtering but not, by historical standards, abominable; the war in Iraq is unpopular, but its social salience pales in comparison to that of Vietnam. One gets the sense that the nation's public ethos during the Bush era just makes people viscerally unhappy and pessimistic in a way that's difficult to quantify or explain.