fix(e2e): replace guest SSH uptime check with LaunchTime assertion

Phase 8.3 polled VPC instance IPs via two-hop SSH. In single-node macvlan
mode the host has no route to VPC subnets, so SSH always timed out (10 min
of dead retries) even when the VMs were healthy (ALB targets passing health
checks).

Replace with a LaunchTime comparison: the daemon resets LaunchTime to
time.Now() when it relaunches QEMU on recovery, so LaunchTime >= REBOOT_START
is a fast, network-topology-independent reboot confirmation.

tests/e2e/run-reboot-e2e.sh

@@ -10,7 +10,7 @@
 #   SSH_KEY  — path to ssh key the runner can use to reach tf-user@WAN_IP
 #
 # Sequence:
-#   1. Pre-reboot: VPC + 2 app EC2s + internet-facing ALB; verify round-robin.
+#   1. Pre-reboot: VPC + 2 app EC2s + internal ALB; verify round-robin.
 #   2. Snapshot pre-reboot state (instance IPs, ALB ENI MAC, ovn-nbctl show).
 #   3. systemctl reboot the host; poll TCP/22 from the runner until SSH returns.
 #   4. Wait for spinifex services + AWS gateway to respond.
@@ -27,6 +27,8 @@ REBOOT_WAIT_SECS="${REBOOT_WAIT_SECS:-300}"
 DAEMON_READY_SECS="${DAEMON_READY_SECS:-180}"
 INSTANCE_RUNNING_SECS="${INSTANCE_RUNNING_SECS:-120}"
 LB_RECOVER_SECS="${LB_RECOVER_SECS:-180}"
+LAUNCH_TIME_SECS="${LAUNCH_TIME_SECS:-30}"
+
 NODE_KEY_PATH="\$HOME/reboot-e2e-key"
 NODE_USERDATA_PATH="\$HOME/reboot-e2e-userdata.txt"