What kind of gradualism?

Matt Yglesias suggests a gradual approach to expanding good single-payer style programs instead of a "big bang" approach to getting universal health insurance now, but based on a cobbled together approach resting on the faulty premises of the employer/individual market systems.

The strongest reason for this approach is that it's good policy. Let's make sure our victories (compromises or not) work, instead of winning passage of a plan that fails and gives all UHC efforts a bad name. But Matt also makes a good political point:

It means the step-by-step defunding of the health-industrial complex that does so much to provide financial support for reactionary politics in America. The alternative -- sticking progressive necks out for the opportunity to direct customers to an insurance industry that hates us -- doesn't make sense.

This approach to fighting big tobacco has worked wonders (as the bunch of recent cig tax hike proposals can attest to). As smoking sections and warnings give way to taxes and bans on smoking in restaurants, bars and public buildings, the number of smokers declines and the tobacco industry has fewer people to whip into a fury and consequently less clout.

One question for Matt though: does this gradual approach have to be straightforward single-payer style programs or is there a role for policies like caps or taxes on administrative costs that make the current system better, rather than building on its flaws (as individual mandates and premium subsidies do)? The taxes or fines from these programs can help pay for the small expansions of medicare and also reduce the insurance industry's profit margins and clout. The drawback is, if these approaches work, they do lower the pressure for change. But is that enough reason to oppose them?