I mostly agree with Ezra Klein's look back today at the stimulus bill and other options available. I completely agree that a $2T bill wasn't possible, and that most of the other suggestions he has collected are some combination of wouldn't have passed, passed anyway, or wouldn't have made much difference.
With one major exception: aid to state and local governments. I'm not absolutely certain that a new automatic stabilizer could have passed, but if Barack Obama had made it a significant priority right away, I think it might well have had a very good chance. And unlike unemployment insurance extensions, which as Klein points out more-or-less happened anyway, additional state and local government aid did not happen, with (in my opinion at least) devastating consequences.
The effects of the recession on state and local governments was quite predictable (although of course the magnitude of the problem wasn't). My impression is that the administration assumed that a bipartisan coalition of governors would be able to successfully push Congress to act. I think that was a reasonable assumption, one that I would have agreed with at the time -- but as it turns out, it was completely wrong.
An automatic stabilizer scheme could be long-term deficit neutral (see for example this idea). There's no real way to know whether a free-standing separate bill could have passed in spring 2009, and it's certainly important to consider the risks of trying -- remember, there was no guarantee that Arlen Specter was going to switch parties or that Ben Nelson wasn't going to, and it's hard from here to know exactly what might have pushed them (or other marginal Senators who voted for the stimulus) the other way. But this is one where, as it turned out, there would have almost certainly been a significant reward from adding it, and as far as I can tell it's totally realistic to think it had a chance.
Of course getting something like that into law (and, for that matter, automatic changes in unemployment insurance) would also have the excellent effect, as Matt Yglesias has been saying roughly forever, of helping out in future down times, too. Which is why it would still be a good idea now, although it's obviously not likely to happen at this point.