First, we'll update that guide to reference newer content in the main SDP Guide: https://swarm.workshop.perforce.com/projects/perforce-software-sdp/view/main/doc/SDP_Guide.Unix.html#_mandatory_vs_non_mandatory_standbys

If a mandatory standby replica is used and you failover to it, even with an unplanned failover, then your won't need to deal resetting downstream replicas. That's the preferred configuration these days.

So, the situations in any action is required on downstream replicas are now reduced, eliminated if mandatory standby replicas are used.

All that said -- To answer your question, I'll need to dig further. If a you failover to one standby replica, and have another downstream standby replica (aka a journalcopy replica), and the replica you failed over to was not a mandatory standby, it is possible resetting would be required. Resetting a standby would require moving the state, statejcopy, and journal.NNN files aside. As noted, with mandatory, that should never be necessary, but an unplanned failover to a non-mandatory HA replica could necessitate downstream resets.

Digression: We are opting not to cover "old school" failover, the way things were done before the introduction of the 'p4 failover' command in 2018.2, which requires standby replicas. Prior to 2018.2, downstream resets were a standard part of failover procedures. Nowdays, the 'p4 failover' method is the only recommended way to go for failover, and in large part the mission of 'p4 failover' was to greatly simplify failover procedures, including reducing the need for downstream resets.