Make status updater production ready

[kate-osborn]

The StatusUpdater component, responsible for updating statuses of Gateway API resources has a number of limitations:

(1) It doesn't understand the leader election. Only the leader must report the statuses of the resources. Otherwise, multiple replicas will step on each other when trying to report statuses for the same resources.

(5) It doesn't retry on failures. This means there is a chance that some resources will not have up-to-do statuses.
Statuses are important part of the Gateway API, so we need to ensure that the Gateway always keep the resources
statuses up-to-date.

FIXME in the code -- https://github.com/nginxinc/nginx-gateway-kubernetes/blob/890fddb787ff3560b9b743647a36b649d498ae51/internal/status/updater.go#L67

Acceptance criteria:

• Address those limitations
• NKG does not print errors when context is cancelled. See Ensure NKG shutdowns gracefully #563

Tasks

Preview Give feedback

1. Horizontal scaling #637
   enhancement refined size/medium +3
   
2. Retry on status update failures #1016
   enhancement refined size/small +3
   

[mpstefan]

Blocked by #637 - We want to tackle leader election before we tackle fixing the status updater.

[ja20222]

This can be closed as discussed on our call today as all items are closed within it
@kate-osborn @mpstefan