On a more practical level: Amazon is gargantuan, and I don’t necessarily trust them to have the scoped vision necessary to run a good medical practice, rather than just a large one that fits their portfolio. This is true for any large company; Amazon is large even by large company standards.
They'll have to be a bit careful with regulation here if they do try it. Clinical trials are where you do that sort of thing.
For the human individual person, the interests (the goal) is to care for their health; but amazon is a corporation: for them the real goal is profit, not health.
I'm saying that for the human the incentives should be all about health. the money, the costs are the "obstacle".
On the other hand, for the corporation the incentives are the profit. The health is a cost (or "obstacle").
with this 'frame' in mind I'm saying the incentives are not aligned on their own because the priorities are in conflict.
What's truly the problem here, is how healthy patients aren't 'good' customers of health-services. Then again, this is not a problem unique to Amazon, but they're getting into this 'rodeo'.
throwing a capitalistic-optimizing 'machine' into health services has been a huge mistake which is seemingly impossible to fix... this is how systems collapse; when the system is so resistant to fixing its 'problems' that only changing the entirety (or a significantly larger chunk) of the system fixes the problems; but the problems have to be 'life or death' (or 'do-or-die') level for this to kick in.
the biggest issue is that for a level of the system there is no problem, people die anyways. but for another level of the system (the human individual perspective) this IS a problem. if/when the system ignores a level of itself it becomes unstable like it's happening now.