I tried writing automation with my finger on the top of the fader, inside the fader indent (where it should be), while touching the screw at the bottom of the fader track, and with a mouse on the screen.  All methods result in a series of dots in the automation curve and all faders on the X-Touch move in short jerky increments, making some noise as they do.  I assumed this was just the way it is.  Is anyone saying they do NOT get these short jerky noisy movements on their X-Touch faders?
 
I haven't been bothered by this because I record in another room, but I do note the unfavorable comparison with how quiet and smooth the equal-sized faders were in the old Tascam FW-1884.  It is also a little annoying while mixing.
 
 

"jerky increments" on automation readout is different of real OP problem (with touch), and it looks like common for all X-Touch owners.
 
Technically speaking, X-Touch is doing what it is asked to do - Sonar IS sending "jerky increments", which at least I can see with faders in Sonar as well. Since X-Touch faders are "fast", they exactly reproduce the "digital world". The issue is rather hard to fix from computer side, and only till some degree. "Smoothing" movements is firmware job, may be with a lot of noise from users, Behringer can update the firmware.
 
Still there was no posts from "real" Mackie unit owners: if they have absolute smooth soundless movements, that is a very good argument to convince Behringer. "Obsolete" Tascam device is less boring for them ;)
 

"The front side" - is that the silver case side?
 

Great trouble shooting guys - it is amazing how an almost throw away comment can spark (excuse the pun) a new line of investigation to get to the root of the issue!

Capacitive touch sensors require capacitive "common earth", for your body and the device. In addition, the capacity from your to this earth and from the device to this earth should be relatively stable. This "earth" is not always (if at all, not sure) related to the ground line. When you use your phone, the "earth" is you. The same happens when you touch the device. Note that the "connection" can be capacitive, so it works not only with touching metal parts but also with touching "in the near".
 
As I wrote before, I am not an expert in this field. The theoretical list of possible reasons: extreme floating the device earth, relative to you (f.e. something inside the power supply); too much isolating or static accumulating environment (table and floor, synthetic and natural covers); interferences from something in the near (some powerful electromagnetic field generator); or (still, sorry) problems with concrete device, which can work better in "better" environment.
 
The environment you can start to check with a phone or tablet (with capacitive screen). Try to use it putting on the table/floor/chair, NOT keeping it in hands but staying as far as possible from the device. Can you notice the reaction is (much/little) worse then usual?
 
You can also try to work with X-Touch in "unusual" position, f.e. sitting near it on the table (not touching the floor). Or taking it on lags. Is there are any changes in responsiveness?
 
All that sounds like a black magic, but when pF range capacity is in game, any voodoo in the near can drastically change the result 
 
PS I am sitting now in the middle of the biggest physic scientific center in Europe... probably if I just go tomorrow outside the building and loudly ask how to deal with that problem, some world class expert in the topic will give an answer. But there is also a big probability there will be more then one... they will be in disagreement... start shouting on each other and I will be accused in organizing all that mess