The Ramesh Ponnuru theory of money

I guess I should comment.
He says:

Another way of putting it is that if the demand for money balances increases, the money supply has to grow to accommodate it.

Not necessarily. If the demand to hold all inventories higher, money and goods, then there is no reason to believe money is being held in higher value than real goods. At the very depth of the crash, early 2009, the desire to hold cash was high. But that period lasted all of eight months, from late 2008 to mid 2009. The number, eight months, is about the settlement time for bankers to figure out what happened.

He quotes Beckworth:

David Beckworth, a conservative economics professor at Texas State University who maintains a blog, has shown that at the height of the financial crisis in late 2008, velocity dropped significantly and the money multiplier fell off a cliff.

No kidding. But it was not just a financial crash, producers crashed also. Transaction velocities drop because that is what the economy intends when it crashes. Producers deliberately intended to go out of business because profit margins evaporated.

Quoting Beckworth again:

That policy [increasing the base but limit lending] helped the banks’ solvency but not the economy. The financial system would almost certainly have benefited more from a broader policy:

If producers could demonstrate a cheaper method to deliver eggs they would have gotten lending. Producers didn't, they had actually crashed because they couldn't find a cheaper method to deliver eggs. Beckworth theory fails. The Fed cannot stabilize nominal spending because money can only flow when the eggs are delivered. Money and goods tend toward coherency, they follow the same path. If resources are not available to deliver eggs then neither will money be delivered.

Bottom line, something real happened to the economy, this was not banking error. Sumner has the same problem.  Ramesh again:

But in mid-2010, the eurozone crisis resulted in a flight to the dollar. Increased demand for dollars again had a contractionary effect, and the Fed took months to respond to it.

No, look again. Oil imports rose dramatically after dropping in 2008, an event that coincided with the federal stimulus (read big trucks consuming gasoline). They were right outside my house, big trucks doing road stuff, consuming oil. Look at oil imports. Compare oil imports to oil prices  which spiked to near $90. The extra demand by Congreoil ss for oil likely delivered a convenient spot to start the Euro crisis. Ramesh has cause and effect reversed.

QE2, though flawed, worked.

No it didn't, look at M1V, once the federal stimulus caused shortages, retail velocity started dropping and it is still dropping. We are still contracting

We have a problem, retail velocity does not have a solution until it drops from 8 to 7.  Beckworth, Sumner and Ramesh all think we live in an infinitely divisible economy. We do not, we live in an economy with a sparse solution set. Uncle Milt makes the same mistake. In the equation MV = PY, the V and the Y are related. Y is log(V). The V is really a series of V(i) where i is a small finite integer. What the economy did is reduce the number of values i, that is how the economy contracts.

What the Fed did is contract, just like the rest of the economy.