Trifield is utilize energy preservation decoding algorithm to redistribute th signal from the LR speaker into an LCR layout. That means, the total energy distribution between 2 and 3 speaker is always the same no matter where the source is panned in the LR soundstage and at what freq.
If they did plain (L+R)/2 or any other non-energy preservation algorithm, a left to right pan of a sound source would sound different with 3 speakers compared to when only 2 are used.
Also, Trifield utilizes wavefront reproduction algorithm so that sound from the 3 speakers works in unison to create a virtual spherical wavefront eminating from a virtual point anywhere between the LR speakers. That is why one need to fine tune Trifield's depth setting so that the wave launching from the 3 speakers's sums together properly.
Steered surround formats like DPLII and Logic 7 send the sound that is closest to whichever speaker is there, thereby, a speaker at a given position frequently would be the sole speaker playing that signal back. This causes individual speakers to become visually visible, but does not require as much fine tuning, since only gain riding are used for imaging between each of the LCR speakers. There is no attempt at wavefront reproduciton, which requires proper matching of the lauch wave and thus fine tuning of the center channel depth setting.