The GUI seems to check the .vmdk file of the last snapshot in the chain to see whether it contains a "ddb.thinProvisioned" entry. Since this entry is only present in the base disk's .vmdk file the GUI assumes a thick provisioned disk.
I'd suggest you delete the snasphot of a single VM at the same time and do this after business hours. Please ensure there's no backup application or some other task which might try to create a snapshot of this machine which the delete process is running!
I wouldn't worry too much about the size of the snapshot or any impact on performance, just be patient and wait for the delete process to finish (which could take some time).
The maximum temporarily required disk space with a single snapshot is approximately the size of the snapshot itself plus some space for a helper snapshot, which contains the VM's changes while deletion is in progress.
André