If they are truly moving more agents to work at home, this could be a good thing for the consumer - if UA does it right.
They can then move to a capacity demand model. In current state, they probably staff the call centers based upon previous year trends and call volumes. What that means though is when IRROPs hits, the call centers get nailed hard and long wait times.
If they move to a work at home model, they could leverage a demand queue where as volumes spike, they can message their employees at home to quickly sign on for 1-2 hours if they are available. Even getting 30-35 agents on for a short period (limited cost to the employee since they can just sign on for a short period without having to get ready for work), could help drive wait times down and avoid a ripple affect.
Would be interesting to see if UA actually leverages this right. I consulted with a company about two years ago to implement and they used to see on average about 45-60 second wait times during normal periods peak at 20-30 minutes during high demand periods, like outages, (which last usually only a short time) to drop to 1-2 minutes during those demands.