To keep confusion from arising, let's get some nomenclature out of the way. "Cirrus Logic" is the name of the chip maker; "Cirrus Extra Surround" (CES) is the name of the processing.

One way to get a handle on CES processing is to understand how it works; the underlying concept is easy to grasp. Most processes that generate a surround-back channel work by grabbing info that is common to the left & right surround channels and sending that mono signal to the surround-back speaker(s). EX, ES and CES all work this way; the difference is that EX/ES decoding goes one step further by cancelling this mono signal from the left & right surround channels.

There are pros and cons to both approaches; rather than sticking with any one process, you'll be better off matching the decoding to the soundtrack. Since CES processing doesn't cancel the back-channel info from the side channels, sounds that are meant to be heard only from behind the listener will in fact be heard from the rear and sides. With EX/ES decoding, those sounds are only heard from behind.

While I haven't lived with a 950, I have listened to CES processing on a several occasions; most recently on a Fosgate FAP-T1 processor. For what it's worth, here are my suggestions:

On 5.1-channel movie soundtracks, I prefer EX/ES decoding. Sounds intended for the rear of the room should come from that direction, and not from sides also. If the surround channels are dual-mono or contain mostly mono content, then I'd use CES; the localization won't be great (tough with mono anyway), but at least the sounds will wrap around you and won't all be steered to the back wall.

On 5.1 music, I've almost always preferred CES decoding over EX/ES because CES seems to provide a more enveloping experience. Yes the rear content is duplicated in the sides, but personally I've never found it objectionable with music because that's a situation where surround ambience is more important to me than surround localization.

Hope this makes sense.

Best,
Sanjay