1. Proposed title of this feature request

Change maxUnavailable on MCPs

2. What is the nature and description of the request.

Modify the MaxUnavailable on all of their clusters. They want to reduce the upgrade timeframe for their ROSA / HCP clusters and they know during their upgrade maintenance windows they may experience downtime and they are ok with that. They'd rather the upgrade finish timely than take 4+ hours.

The customer would like to 

3. Why does the customer need this? (List the business requirements here)

We have been setting maxUnavailable to 1/3 of total worker count for several years now, to expedite our upgrades as we only have a limited outage time window for performing upgrades. We want to set this for our rosa clusters as well. The last rosa upgrade from 4.15.48 to 4.15.52 took approximately 4-5 hours to complete, for a minor upgrade we've never seen it take more than 1-2 hours in OpenShift UPI.

Upgrade durations that extend beyond our available outage time, resulting in outages being delayed. An upgrade timeframe that is similiar to our UPI clusters (1-2 hrs) is desired.

4. List any affected packages or components.