Market monetarism, as propagated most prominently by Scott Sumner's (excellent) blog The Money Illusion , argues that recessions come about due to a collapse in demand. This is a problem because prices cannot adjust downwards quickly. Instead of a costly adjustment period we can simply boost demand by announcing a target and credibly committing to do the necessary quantitative easing (buying gilts to inject money in the system) to achieve that target.
This makes a lot of sense. Markets are finding it hard to clear; we boost AD to put the situation back where it was; now markets find it easier to clear. But lots of the best market monetarists, including Scott, Lars Christensen  and many others, argue that right now what we need is more stimulus, because the economy is still in a bad shape, and it is still due to a shortfall of demand.
Last Tuesday Professor George Selgin  delivered an extremely interesting lecture at the Adam Smith Institute making the case for productivity driven deflation. He said he agreed with the market monetarists that there is "bad deflation"—the sort that means nominal rigidities stop markets from clearing—but there is also "good deflation", from productivity improvements—and this is not associated with unemployment, stagnant or falling GDP, or any other cyclical issue.
After the talk I quizzed him on whether he agreed with the market monetarists that even though the ideal is a rule-based system, as opposed to the current discretionary way policy is set , right now the best discretionary policy is more easing, because that's probably what the ideal rule would require.
Prof. Selgin disagreed, arguing that we didn't need easier policy, and if you look at the graph above there's at least apparent reason to agree with him. Nominal GDP—aggregate demand—is not only well above its pre-recession peak in the US, but is growing at an apparently steady rate, roughly in line with its long-term trend. If the high unemployment in the US is down to insufficient demand combined with nominal rigidities then why hasn't a long period of higher-than-pre-crisis demand brought unemployment back down.
According to Selgin, policy uncertainty and pro-cyclical strictness in enforcing regulations (particularly risk-weighted lending rules that rate Greek bonds as zero but loans to small business at 100%) are holding firms back from investing their cash piles in capital and it is this that is stopping the robust recovery. He made the point very convincingly and despite trying hard to argue against it I couldn't find a good reason to disagree, except that I hadn't seen a good measure of the importance of these two factors so it was hard for me to compute how big their influence really was.
But many market monetarists—along with New Keynesians and most others—seem very sure that insufficient demand is the overriding factor holding back recovery, in the US as much as the EU, UK and Japan (where NGDP growth is further below trend). So my very genuine question is: upon what arguments and/or evidence do they rest this belief?