The front rear reversal is a phenomenon in which we have trouble localizing sounds that are directly along the midline. We rely on several cues to determine location including level differences and arrival time differences between left and right ears. These are obviously absent when a mono source is along the midline. That leaves the way our outer ears modify incoming soundwaves dependent on the angle at which they approach our ears, commonly referred to as hrtf, or head related transfer function. People with complete loss of hearing on one side can still localize sounds fairly well relying on only this cue (obviously much better on the side with hearing intact). Strangely enough, with sounds along the midline and other cues gone, hrtf can fail us as well making it difficult to figure out if a sound is directly ahead of or behind us. This happens more frequently when there are visual cues in front, like a huge screen filled with action that our brains are happy to attribute as the source. We also struggle more when the source is above ear level (I suppose because in an evolutionary context things which ate us tended to be at or below eye level, not above). For this reason we can be fooled into thinking a high mounted center is actually at eye level more easily than one below eye level, which is why I and many others typically recommend a center be mounted above the screen rather than below if the choice must be made. And it also makes front rear reversal more likely when the rear center is mounted higher on a wall, which it often is. And this is why lexicon and others wishing to implement overhead effects don't have an actual overhead speaker in their configurations.
This was the result of research into surround sound methods and is why the industry quickly moved from 6.1 to 7.1 format (and why some of the companies more attuned to the research, like lexicon, skipped 6.1 entirely). This research showed that if two speakers separated far enough (it doesn't take much) reproduced the rear signal, the change in incident angle and change in hrtf was enough to stabilize the rear image. Strangely enough I don't recall where all I've run across this information. Probably in some Dolby labs white papers, aes articles, and likely many other places, though I can't recall any specific sources off hand nor does a ten second google attempt locate the obvious original sources.
In any case, this is all a theoretical explanation. While the effect is real, has been well documented, and was compelling enough to steer the industry away from a single rear channel, it doesn't mean any particular person will necessarily have a problem in his room. There may be room asymmetry, you may not sit quite on the midline, there may be furniture or other nearby objects in the rear of the room that produce strong localizable reflections, a horizontally configured center channel may have two midrange drivers separated just far enough, your rear center may be mounted relatively low, or a host of other reasons may contribute to minimize the problems in any given application.
For this reason, you can always try it out. The danger is that you simply won't know when the phenomenon occurs; you may hear a sound coming from the front that was actually produced from your rear center, and not even realize that sound effect was intended to come from the rear. This could happen frequently and you still may be oblivious. And perhaps that wouldn't bother you, even if you figured it was happening. But since this is a known issue, and it was enough reason for the entire industry to move to a pair of speakers in the rear, I still feel justified in giving the initial advice I did... only use a single speaker for the rear if you have to.
