They are hard to see unless you find the right photos. I noticed that the taper at the front of the base of the pylons are different lengths. There is a slightly different port arrangement between the sides. The starboard side has a few more ports (on the removable panel that covers the mounting point). The hatches on the bottom of the saucer are uneven. Two pull down and then the port side airlock doors are removable to reveal the airlock. The only other things I've noticed are some small variations from side to side.
Except for one kitbash that has not be identified as screen used, the only uses of a single (or third) warp nacelle are the Klingon Bird of Prey, The alternate future 1701-D from All Good Things, and the Kelvin from the 2009 movie. The kitbash looked like this (the drawing is a bit clearer than the photos of the model - the model has no registry or name)
http://www.geocities.ws/cpt_kyle_amasov/Medusawn.jpg. Single Nacelles or triple Nacelles originate with Franz Joseph's Technical Manual. We have seen the drawings on screen of his scout/destroyer design and a couple of his names and registries were used in TMP (including one Dreadnaught). But in general Star Trek has confined the Federation design (and most others as well) to twin or quad nacelles.
The Klingon Bird of Prey is an oddball in this conversation because it has no obvious nacelles. It has the large impulse engine, two wings with weapons at the ends, but no nacelles or pylons like the other ships. The Oberth class is also odd because the warp engines have the saucer between them. Both of those designs violate what Mr. Probert has shared as Roddenberry's rules of Starship design. Actually, the Defiant is similarly odd.
The three rules that concern warp nacelles (the fourth is that the bridge is on top... Shenzhou broke that one) are that the nacelles must be in pairs, the nacelles must have at least 50% line of sight, and must be fully visible from the front. Well, that last one is blown out by the 1701 Refit and the Excelsior. And Voyager breaks that same one when at warp and the 50% line of sight when not at warp (and I'm not sure that it quite achieves 50% when they move up to the warp position). The even numbers was not broken for a long time, save for the displays of FJ's single nacelle configuration. As far as I know the alternate future 1701-D is the only one we have seen. The Oberth class breaks the 50% line of sight rule, as the the Phoenix. There is no line of sight at all. The Klingon Bird of Prey has no obvious warp drive so it all must be internal. The merchant ship built for Star Trek III also has no obvious warp drive. The TOS Tholian ship seem to either be a single or triple depending on how you break apart the design. So the rules have been frequently broken and I see no reason to follow any of them. And there is nothing about the canon nacelle designs that preclude variants that are single or triple. As we see in the Constellation Class which has a near identical nacelle to the Constitution and Reliant classes. But also, the Federation design seems to indicate that the twin or quad configurations are more suited. It could be like the number of rotors on a helicopter. With two rotors, it has no need of a tail rotor. So there may be some system in single or triple nacelle ships that corrects for the difference and allows them to work. But some other designs don't have the split nacelle warp drive system. The Vulcan system we see is completely different. They were using the ring system. So confining all ships to Roddenberry's rules is impossible because too many canon designs don't follow it exactly. I would say that Roddenberry's 3 rules for the nacelles would be what the early engineers found to work best and what most ships follow, but breaking them is fine by the 23rd century because the technology has advanced enough to stretch beyond those confines.
