The problem with using capsman is that checking config locally doesn't actually have to reflect running values. One thing that CAPsMAN definitely doesn't do is overwrite configuration stored on CAP devices. So running export doesn't show any of CAPsMAN-provisioned settings. Running "monitor" command (in sections where it exists) might show actual state of "monitored" interface ... but it might not show anything for CAPsMAN-controlled wireless interfaces. In wifi CAPsMAN environment, running monitor against cap interface shows running values:
# on CAP device
/interface/wifi/monitor [ find where name=wifi1 ] once
;;; managed by CAPsMAN BA:69:F4:20:A5:49%vlan-99, traffic
;;; processing on CAP
;;; mode: AP, SSID: mkxTestwAP, channel: 2412/ax
state: managed-by-capsman
# on CAPsMAN device:
/interface/wifi/monitor [ find where name=cap-wifi1 ] once
state: running
channel: 2412/ax
registered-peers: 0
authorized-peers: 0
tx-power: 23
channel-priorities: 0:2412/ax,0:2432/ax