Having supported NetApp on FAS270/940/960 and up with Ontap 7.1 and up....as well as EMC N70, NS702, and NS740 Celerras on DART 5.3, 5.4, and 5.5, I will gladly continue to work on NetApp gear.
From a CIFS standpoint I don't even see a comparison. Having to use the MMC to manage any and all CIFS on an EMC datamover is severly limited, especially when you are having CIFS issues.
On all the versions of DART I worked on, if the checkpoint file system hit 100% because a user put a ton of data on a volume, then deleted it once caught, you have to delete all the checkpoints and configure it again. Or management tells you to use autogrow and the thing gets to be TBs in size.
However on NetApp, unless you use autodelete, snapshots can impact production operations, which can be an even bigger problem than losing checkpoints.
Having the datamover set behind the control station and only working through it leads to complexity that has little value in my opinion. Having fs_, nas_, ckpt_ and other command structures simply does not feel efficient. Constantly movnig back and forth between hardware/software/system commands is grating after a while.
Active/standby node setup on EMC is nice though, it prevents managers who just don't get why an active/active cluster is not at 100% from killing you when you have a failure.
And if you need mixed mode access(Unix accessing Windows or vice versa), without AD and LDAP/NIS integration, EMCs permissions are easier to manage as each file has Unix and Windows metadata. In today's world it shouldn't be an issue, but it is in far too many places.
Your mileage will vary, no two situations are ever identical. I simply have found NetApp to be a much simpler, but no less robust system in my dealings.