Sync operation unnecessary waits for a healthy state of all resources

[alexmt]

Describe the bug
Sync operation unnecessary waits for a healthy state of all resources even if user has a single "wave"

To Reproduce

1. Sync any deployment which consist of a single wave.
2. Sync operation stays in running state until all resources are healty or degraded.

Expected behavior
The last waive should not wait for a healthy state of all resources. This is changing sync operation behavior.

Version: 1.1.0 (unreleased)

[alexec]

We should brainstorm this.

Sync waves change how the op status is determined.

At any point during a sync op, including post-sync, we can end up with a situation where we cannot continue (everything is degraded) or progressing (e.g a pod in crash loop backoff). We don't' differentiate recoverable (crashloopbackoff) from unrecoverable (invalidimagename), it means we can't tell if a degraded resource will become healthy. Because these situations, therefore, may resolve, the wave could become healthy, and the op continues.

This could happen a long time after the sync is requested. (By "long" I mean, there'' Git commit that changes the app.

This op will remain in progress until it is terminated: