Does iffy (but press-worthy) science pay?

Astronomers get far more press, per capita, than our bio & chem colleagues; Marla Geha points out that the 1E4 astronomers worldwide get roughly the same press coverage as the few 1E5 chemists and biologists. Presumably this is because astronomy excites & captures people's imaginations. It also helps that startling results can often be explained in a few words ("We've found 200 planets orbiting other stars"; "We see a meter-thick ice layer on much of Mars"; "We've seen back to only 300 Million years after the Big Bang".)

It's worth taking a moment to appreciate that public support. Since my research has no practical use, the reason I get paid to do it is, simply, that the public thinks astronomy is cool.

Unfortunately, reporters often lack the expertise to question the press releases they're fed. As a result, gee-wiz but technically faulty results find their way into newspapers, APOD, and popular science magazines. Since the press and the public have short attention spans, these results don't get retracted.

Thus, occasional papers report very surprising results, and get loads of publicity -- but it's pretty clear to colleagues that there may be a more mundane explanation, or the result's total crap (systematic effects, miscalibration, etc.) Did the authors mean well, but fail to follow Feynmann's famous dictate, that you must not fool yourself, and you are the easiest person to fool? Or do they bargain cynically that a splashy-but-crap Nature paper is worth more than a cautious-but-prosaic ApJ? And if so, are they right?