Daniel Gross thinks the Democrats may do better in 2010 than you think. But he has a caveat:

Yale economist Ray Fair has developed a relatively simple and highly accurate economic model for predicting what percentage of the total vote parties will receive in presidential and congressional elections. It relies on inflation and economic growth over the seven quarters before the election. Fair, who has a somewhat optimistic take on growth for 2010, says Democrats are likely to get a 51.6 percent voting share this fall. "The Democrats are not going to gain seats, but it doesn't look like there will be anything like a disaster."
One caveat: high unemployment and housing—factors not considered in Fair's model—may play an outsize role in people's feelings about the economy.

I think this caveat wipes out much of the power of his argument. In normal times, economic growth is a decent proxy for "economic well-being." But when you're climbing out of the worst economic crisis since the Depression, a normal rate of growth isn't going to feel like prosperity. Moreover, the last two economic expansions featured excruciatingly slow job growth at the outset. And the last recovery basically saw no wage gains at all. That is, the historic link between overall economic growth and prosperity for most people has been severed. So I have a hard time seeing the economy being anything but a huge drag for Democrats -- and, consequently, major Republican gains.