Context matters. Knowing what we did back then, Republicans had just shut down the government and were en route to an electoral beating the following year. Thus, Flo's prediction was perfectly reasonable while JerryArkansas's was hackish, since he was naive enough to predict that ephemeral political conditions would change within the year and create a good environment for the party out of power.
How do you know it was naive and not prescient?
Midterms often go against the President's party. The map may not have felt right at the time, but that's because we naturally have a bias to recent events. Now we know the shutdown didn't matter much in the end. But one could have reasonably forseen that at least as a possibility a year ago too.
