the first one should refer to the timer to wait before treating the current querier (the router that sends out on the vlan IGMP queries every 60 seconds) as down.
for the second: the router tracks for each multicast group the ip address of the last host that has sent an IGMP report ( it doesn't track all the hosts also because igmp reports are suppressed by hosts).
if an IGMP leave is received for a group from last reporter the router wants to check if there is somebody else interested in that multicast group so it should send out a query for that group.