All this is subject to the usual caveats about markets only telling us about effective demand (of people that have money), and nor reflecting any changes in social preferences that might come about through public deliberation, as well as the usual public goods problems and externalities. But any source of information has its defects, unless you're God --
[WS:] That is the bounded rationality/transaction cost argument. Basically, finding information can be costly, hence many people "satisfice" i.e. settle for the first acceptable ("good enough") alternative they encounter. Therefore the "signals" they are sending by their purchasing decisions are not necessarily indicative of their "true" preferences provided that they had "full information," or alternatively, provided that acquiring "full information" had zero transaction cost.
The transaction cost economics (TCE) folk take that a step further by proposing that transaction cost of acquiring information can be reduced (if not eliminated) by "hierarchies," by which they mean organizational structures (governments, corporations, etc.). Therefore, organizations are more efficient than independent free market actors, because ceteris paribus they have better information (or information at lower marginal cost).
In plain English, a health care organization is in a better position to find information about the efficiency (value/cost ratio) of health care providers than individual patients. It makes sense on its face.
However, the critics argue that "hierarchies" may reduce transaction cost of information, but introduce other transaction costs (such as crooked officers of the organization that embezzle resources). So if the "hierarchies" do replace "markets" it is not because of their greater efficiency, as the TCE crowd claims, but because of their size, monopoly position, and power. In plain English, governments or large insurance companies can use their power (regulatory or economic) to force the suppliers to lower their prices without reducing the quality of the goodies.
I think that both sides may be up to something. I like TCE arguments because they make a good case for central planning.
Wojtek