The health check microflow is a 'special' microflow in that it's accessible via the M2EE API. This is a an administrative API that is used by tools such as m2ee-tools and the cloud portal. It is normally never called, but:
The idea of the health check microflow is that you implement it yourself and check if your application is in a healthy status. Healthy is something you'll have to define yourself for that specific application. It's basically a tool allowing you to monitor how your application is doing, if you choose to implement it. For example you could check how many scheduled events have the status 'Running'.
I'm not sure how/when this microflow is called in the various deployment environments, maybe someone else can shine a light on that?