I would absolutely love to see something like this happen but with challenges from the major players in the healthcare market: Philips, Siemens, GE, etc. There is a spectacular array of startups and applications that could spring up if the extremely thick and crusty shell is peeled off of healthcare data access.
I worked on a team that spent 5 years on a realtime data protocol for one of these companies and didn't get very far. I left 6 months ago and they're still working on it.
Right there with you. Practice Fusion, while a popular vendor, will never be able to get at the major providers with their "free" EMR. However, their API is a step in the right direction.
We are working on technology which begins to get at this locked-in data via an API package which would be installed at individual institutions. The technology utilizes HL7 messaging to get at data from the major players (ie. Epic, GE, Cerner). The problem is the main software vendor, Epic, has built their technology up, by erecting proprietary barriers, an API would kill their entire model. Therefore the best way to get at this industry is through the providers, as they are becoming increasingly burden by these proprietary, legacy systems.
All of these developer challenges popping up remind me of something we've long loathed in the advertising industry: spec work aka contests. It's worth noting that compliancy is a billion-dollar problem, and if you help them solve it you get $2,500 and a ticket to Health 2.0.
From a high level perspective, "health 2.0" is very interesting to me, and the research I've done in recent months has led me to believe that it's not a place for your run of the mill startup. You need some serious gunpowder to survive. However, it does seem ripe for a few disruptions.
I'm currently working with Health 2.0 to access these technologies. In my opinion the most exciting of these is the challenge issued by Practice Fusion. This is the first time I have seen a vendor begin to open up their system.
As was stated in a previous comment, there are a large variety of applications which would benefit from access to actual medical data, and not require manuel entry. IMHO, this manual entry is why services like Google Health and Microsoft HealthVault will never take off.
I worked on a team that spent 5 years on a realtime data protocol for one of these companies and didn't get very far. I left 6 months ago and they're still working on it.