To add to Larry's excellent discussion, I would like to point out that  
the significance of the neutrality of money assumption concerns policy.  
Throughout the 50s and 60s, in the heyday of optimism about the  
potential effectiveness of discretionary monetary policy, the theory  
associated with what Mark Blaug called the Cantillon effect was  
available to typical economics students only in the history of thought  
texts (e.g., Economic Theory in Retrospect).  A generation or two of  
macroeconomics students were trained to disregard the Cantillon effect.  
Economists who were critical of this optimism, like Friedman and  
Buchanan, focused not on the disruptive injection effects of an increase  
or decrease in the quantity of money but on the political processes that  
pressure the actual monetary policy decisions. Both emphasized the  
dangers of discretion in a democratic society. Only those precious few  
who ventured into Austrian economics or into the history of thought  
learned something about the micro-economics that was most relevant to  
evaluating discretionary monetary policy.  That is, only they learned  
about the Austrian trade cycle theory. 
 
Also to elaborate on Larry's point about the particular means of  
increasing the quantity of money (namely, through the loan market), the  
key to understanding the Austrian argument is to recognize the demand  
for present goods in relation to the demand for future goods. Mainstream  
macroeconomics models used a concept of money saving that almost  
entirely disembodied this demand. By doing this, it diverted attention  
away from Mises's main idea -- that unexpected money injected into the  
economy through loan markets distorts the signals that consumer-savers  
give to entrepreneurs about their near future goods demand. This  
diversion was part of the Keynesian legacy. 
 
Pat Gunning