4.x: Add metrics for concurrency limits and fault tolerance

[tomas-langer]

Currently we do not have metrics in fault tolerance components, and we did not introduce them to concurrency limits (https://github.com/helidon-io/helidon/issues/8897).

We should find a good way to be able to expose information from these components that can be seen in metrics. We should honor the name of the component, to allow differentiation between instances (i.e. we can have a fixed metric for a client and a server, and these should not be seen in metrics as the same instance) - this can be achieved through name, which can be specified by user in configuration, or when explicitly setting an instance.