the database gods still won't me comment here, so i'll try it here...
medicare admin -- depends how you want to calculate it.
from the latest medicare trustees report [these are somewhat interim numbers, i'm informed, but they're not usually too far off from what cms calculates later in the year], pg 11: in 2008 medicare took in $480.8 billion, paid $461.6 billion in benefits [doctors, hospitals, drugs, etc], and paid $6.5 billion in admin costs. that's 96% going to care, 1.4% going to admin costs, and 2.6% staying in the trust fund[s].
distortions: how much is crap like this costing us, all told? you could get a phd studying just that one question. and the 'practicing medicine without a license' aspect of it... grrrrrr. i worked for a few years as a middle-person: buying and selling, negotiating contracts between me and vendors, me and subcontractors, me and customers/clients [from sole proprietors on up to multinationals], in wide-open, not-very-regulated fields -- if it's an atmosphere you've never worked in, it's almost impossible to imagine all the permutations. i could go on at length, but will refrain from inflicting that on you. :)
fee-for-service: canada, japan, france, all have lots of ffs in their systems, and lots of solo docs or small group practices, probably more than we do. the difference is that they charge so much less for each service that you can have lots and lots of services for an overall reasonable cost. i don't know as much about other countries [i think germany has more managed care / large group practices like the us does, but don't know for sure].
more on payment methods from steffie woolhandler and david himmelstein.
i get what you're saying about the provider side, but if insurance companies are either jacking around the providers, or colluding with them, on prices and payments, then you no way can you estimate the impact of trying to take a chunk out of the providers' hides to pay for reforms. put a choke collar and short leash on the insurers first [and none of the proposed bills do this, so far, in this regard], and then maybe you can talk about 'incentivizing' physicians to do 'better' [although the vast majority of doctors go into medicine because they want to help people, so they're not really the ones who need 'incentives'].
some stuff about france's system [i would dearly love to have their system, but i'll happily settle for canada's].