Oh, I know very well about meaningful use and the incentives thereto. I give lectures on this subject, albeit usually in the context of the risk management implications of EMR.
Ostensibly, one of the primary rationales behind this type of information sharing is to use data mining of de-identified data in large populations to help determine the most safe, useful and cost-effective means of treatment for a given condition, and reduce the wide variability in treatment. I personally am in favor of this from the standpoint of using evidence-based medicine more widely.
I would point out that the insurance companies have already been doing this for a couple of decades, and I have far more concern over the privacy implications of Blue Cross getting this data than I do the government. Especially as genetic testing becomes more common, and we can do a better job of identifying your chances of getting a certain disease in the future, will the insurance companies use this information to deny you coverage or charge you more? If I, through no fault of my own, have a genetic predisposition towards, say, diabetes, should I pay more for my healthcare coverage?
If you ever look at the paperwork for signing up with your insurance company, you are already giving permission for them to access pretty much anything they want in your health records, that is, if you want them to pay for it. The Medical Insurance Bureau has been collecting and storing information about all of us for quite a while now.
http://www.privacyrights.org/fs/fs8-med.htm is an interesting read on medical privacy issues. People here would be astonished as to how many fingers are in your medical records.