I suspect the issue was a result of vm monitoring ha action in that some vm's were restarted because of vm monitoring feature of HA thought the guest had stalled so rebooted it but thats only a guess
In case its useful to others
I removed the disks, obviously keeping the default of not deleting the disks. Just because it says hard disk 2 doesn't mean its an appliance disk as the disks appear to be listed by SCSI channel order and when you re-configure the appliance the hard disk 17 in the image might have become hard disk 3 but without changing the vmdk name. The guest disks were listed as independent (not snapshot-able) on the appliance but were snapshot able on the guests.
After the reconfigure of the appliance completed successfully the guest which had disks attached don't appear to need a consolidation according to vSphere UI and when I edit settings the disks on the guests don't have the obvious snapshot look of diskname-0000000x.vmdk and using RV Tools there doesn't appear to be any zombie disks where the snapshot disks should be. There also were no new storage related events or tasks for the guest vm's involved after reconfiguring the appliance