cooperative_monitoring secondary MaxScale can ignore failover by primary MaxScale

Primary A, replicas B & C. Replica B goes down. Then later, A goes down. MXS1 does failover. MXS2 does not abandon A as it still sees B replicating from it (using old data).
Addition: Not a real bug, as the situation persists only for failcount monitor intervals. Assuming both MaxScales are configured similarly, MXS2 should abandon A around the same time as MXS1 performs failover.