I'm reluctant to jump into a long discussion of the specifics of libertarian public policy -- mind killer and all that -- but in light of the terrible account of itself libertarianism has given you and SoullessAutomaton, maybe a few nonspecific comments are in order.

There's such a thing as libertarian public policy research. It happens in think tanks. It gets done by academics (mostly economists), it incorporates peer review, and it usually doesn't hold with the kind of boorish behavior you're describing. Many of its hypotheticals are imports from the most inconvenient possible world. Specifically, it acknowledges that market failures exist and that government intervention is sometimes the most effective way to deal with them; that regulation has legitimate uses in service of the public good; and above all that pragmatism and compromise are the only virtues that can survive in the political arena.

Like most public policy it is essentially utilitarian, and its specific claims center around the idea that society is too complex for any central authority to administer efficiently. That's to say, while there are many good ends the government might achieve through intervention in the economy or the private lives of its citizens, the costs of such intervention -- money spent, conventions altered, expectations shifted, power grabbed, responsibility abdicated, and goals co-opted by the political process -- are rarely less than the benefits.

You may take issue with any of these claims, but hopefully you can agree that the framework I'm developing here supports more sophisticated answers to the question "As a society, what should we do?" than just chanting "Private GOOD! Public BAD!"

In the specific case of your test of Regulatory Scheme X, the thoughtful libertarian position might go something like this:

Accountability is great. Empirical validation is great. But in this case, your test is a non-starter. No one is going to want this to happen. Drug companies will resist the removal of their products from the marketplace. Doctors will see the legislative call to dispense with certain treatments as a threat to their professional autonomy. Crossover between the AMA and FDA will favor an equilibrium where most experts already support the status quo. Insurance companies will use the opportunity to demand changes elsewhere in their payment structure. Any one of these groups can scuttle the whole project and throw your whole party out of power in the next election by letting it slip to the AARP that you're planning on taking away something previously covered by Medicare. And even if you manage a legislative or executive miracle, anyone in Area A who wants the banned treatments can just migrate to Area B and get them there.

None of this should be interpreted to rule out the possibility that the test itself could yield invaluable information, saving lives or huge amounts of taxpayer money. But no regulator or legislator has any direct incentive to risk his career and all his political capital for nobody in particular. When talking about cost-benefit analysis, it's important to remember that government officials implicitly measure costs and benefits to themselves, and that many of the responsibilities government arrogates to itself go unmet as a result.