I think the idea they were going for is that Google Health would be an EMR intended to be used by institutions except that the patient is in control of their record - enabling them to take their health record to any provider. The problem is that institutions have close to no reasons for investing into something like that.

I think Microsoft HealthVault was more of what you describe, where it's a portal intended to be used by patients.

It turns out re. docs and nurses, not really. It's important to understand the who pays for what in healthcare, and how many links in the chain exist between payer, payee and who benefits. Physicians and nurses are not able to pick whatever software they want to view patient data because 1.) you need high security all around that data, and 2.) that data is primarily useful to someone else. It's useful to the doctor because they want to treat the patient. It's a lot more useful to the hospital because they want to get paid for what was done.

So the hospital is willing to shell out a ton of cash so that they have a Very Secure system that helps them better manage their cash flow. If you walk in as a doc with (for example) a nifty phone app that lets you track your patients in real time, you'll confront 2 problems: 1.) the hospital won't let you connect because it'll assume (possibly incorrectly) that it's a huge security hole. 2.) even if they did let you connect, their EMR is non-standard (basically all of them end up non-standard) and there's no lingua franca protocol for medical data, so your app is incompatible.

It's really really tricky to solve all those issues AND have it be cheap enough for a doc to be willing to shell out for it out of pocket. I personally you'd need that kind of consumer-side scale to get enough competition to get medical software to innovate more quickly. I will say a common interchange format would help a lot, that's one of the objectives FHIR had, and the industry has been slowly adopting it.

You accurately point out revenue personnel and practice managers as key players here, and yes they're the ones that usually push EMRs etc. But the front line workers are kind of disconnected from the design and functionality of the EMR.

Put another way you have software that's bought by one person (hospital/provider), it's used to collect data by a totally different person with different needs (docs/nurses), and it's paid for eventually by another person who has no idea how it works nor any direct pressing need for it (patient/but let's be real, insurance company/government).

Without a tight connection between who pays for the software, who uses the software and who benefits from the software (most consumer software is a pretty tight loop), it's hard to have a good virtuous cycle of innovation coming from customer/user insight/observation/needs.