First, is the statement true?

So why does the banking crisis in 2008 cause low NGDP in 2011?Scott and Arnold ponder

The 2008 banking crisis is not causing low NGDP growth in 2011. Something else happened in 2008 besides a financial crisis.

Even so, they did have the power to prevent NGDP from falling 10% below trend in 2011, Ben Bernanke just reiterated that point yesterday.Scott

Ben and Uncle Milt were both wrong, the central bank cannot cause general inflation, at least not for 3 years and definitely not when material shortages plague the economy.

I’m simply doing mainstream economics, right out of all the textbooks. 

There is a mistake.

In the future we should just let non-economists write our textbooks; at least if we plan on adopting seat-of-the-pants, common sense views as the OFFICIAL DOGMA of macroeconomics.

No, just change the measurement norm too represent seat-of-the-pants common sense. What makes depression era economics different is the finiteness of the economy cannot be ignored, smooth approximations only work during Spontaneous Order. Corridor Stability is in effect during normal times.

In between corridors, the Fed had few paths forward, the few things it juggles don't appear smooth to the first decimal point. The Fed has to play a game called patterns of sustainable trade, what things can it do repeatedly.

Another hypothesis.  We are in a financial crisis because we notice the economic skew, we notice that money is not smooth.

What happened to make us more observant? Information technology, we can track real goods better than the banks can track  money.  So centralized financial systems are falling further behind the information flow, late to the party.

Solution? Digital money so the bankers curve can be constructed from real goods curves over any arbitrary region.  Then the NGDP  distribution (over time and space)  will always map to  real goods distribution, as currency markets can update the 'in pocket' value of a digital hard coin.  The automated currency web bot becomes a Huffman encoder, look for unattended aggregates of  various goods types, money included. Then the web bots attempts the purchas of an unattended aggregate of some good, over time and space, finding the best mutual entropy match of another traders.  You get the spontaneous order corridor of money traders, money holders and real goods.

How would the robot traders work
Well, I suppose the web bot starts searching for bonafide lists of things on the web.  It maintains a set of [-iLog(i)] for various i.  Then it looks for a scale, a median value of that thing in terms of the median value of something else.  Given that, the web bot can go look for trade partners, who have their own lists of discovered aggregates.   Traders actually do this all the time, they are entropy encoders
I have a handy Huffman encoder on the right, works for spread lists of economic data.  It will naturally find aggregate of -iLog(i) that are unattended.