Another possibility is they thought it was a cool idea, then found out how regulated it was, how it would be nearly impossible to add any features without massive government oversight, and got out.
I was an intern on the Google Health team and witnessed many problems unrelated to external barriers.
1) The original codebase was a nightmarish mess of Java/GWT code that did very little (my first real-life encounter with a FactoryFactory). Most of the developers from this first version drifted away from the project and by the time I arrived there was a second team who were talented, but, unfortunately spent much of their time slowly refactoring other people's crappy code.
2) I didn't observe very much clear product vision. Instead, we had a paranoid obsession with matching the features of Microsoft Health Vault (which was equally meandering & useless).
3) There was a huge top-down pressure from Marissa M and other high level managers to make Google Health into something astoundingly successful, suffocating any possibility of incremental progress and disempowering the actual developers.
4) We had one MD on staff and very little other experience with medicine or healthcare. The developers were very far from the problem domain and relying on a game of managerial telephone to ascertain what the current state of medical record management is and what improvements are possible.
Anyway, in short, my experience was that Google screwed up.
Another possibility is they thought it was a cool idea, then found out how regulated it was, how it would be nearly impossible to add any features without massive government oversight, and got out.