That's good, although you keep showing but one interface. The reason I keep trying to confirm you've reconfigured all HSRP interfaces is this little nugget in the reference I earlier provided:
The timers configured on the active device always override any other timer settings. All devices in a Hot Standby group should use the same timer values.
Assuming your earlier reduced timer test, which appeared to not correctly work, was in fact correctly configured, interactions with emulator can cause odd issues. I.e. different code paths might be involved. Since you're now using millisecond timers, then the following might apply:
HSRP BFD Peering
The HSRP BFD Peering feature introduces Bidirectional Forwarding Detection (BFD) in the Hot Standby Router Protocol (HSRP) group member health monitoring system. HSRP supports BFD as a part of the HSRP group member health monitoring system. Without BFD, HSRP runs as a process in a multiprocess system and cannot be guaranteed to be scheduled in time to service large numbers of groups with hello and hold timers, in milliseconds. BFD runs as a pseudopreemptive process and can therefore be guaranteed to run when required. Only one BFD session between two devices can provide early failover notification for multiple HSRP groups.
This feature is enabled by default.
Interesting too, is @MHM Cisco World's mention you need to use BFD. (In my past experience, I recall using HSRPv2 and millisecond timers before BFD was supported on many platforms, and subsecond switchover was often impractical.)
