Why would an application's performance degrade after failing over it's database on a SQL cluster to the passive node?

Question

When repeatable maintenance is performed on a SQL cluster (which has over 80 application databases hosted on it), we fail over to the passive node, and patch the active node, th