[ExtDNS] Flexibility to set "log-level" option

The OpenShift External DNS Operator version 1.0.0 passes a hard-coded "debug" option to the "--log-level" parameter when starting an external DNS container (ref in GitHub).

While this option is perfectly fine for a development/lab environment, it might be a nuisance for higher environments (pre-prod/prod) since it generates a lot of messages in the logs which might generate a lot of data, and cause a very large volume of traffic when sending the logs externally to the cluster (like Splunk or Elastic Search), specially in large clusters or where there is a lot of applications being created/destroyed on a regular basis.

Ideally, the operator should allow the cluster administrator to customize the log level to one that matches their interest in any given environment.