It seems inconsistent with walrasian pure and perfect competition, where profits tend to zero and the optimal outcomes are achieved.
Eliminating the inefficiencies, ie profit in pure and perfect competition, is the way the system works - non null profits attract new businesses. That's basic economics, with very special edge cases (ex - natural monopolices)
I'm not sure exactly why or how economy as a whole would do better with more inefficiencies at each point, and money moving more as you suggest.
Are you referring to the equation of the velocity of money (MV=PQ) ?
It says nothing about how desirable the outcome is (ex: broken window fallacy)
It also has to deal with the dynamic disequilibrium that is caused by entrepreneurship and innovation, which "creatively destroys" previous markets in favor of new ones. It is this process that is the source of all profit and economic growth, at least according to Schumpeter.
Does anybody really want that?
The more times money moves around the more money others stand to gain. The economy as a whole does well when moving between parties at a consistent rate.
Eliminating the inefficient transaction doesn't just hurt profit. It hurts the very core of how our economic models currently work.
That being said Bitcoin is currently no threat to that system. It's hard to imagine it ever being honestly.