Very cool! So you are getting better results from using the 4 corners over the 4 mid points? I had thought mid point positioning was considered the gold standard by Harman et al. over four corner placement...
Most also manage to skim over or forget the assumption made in the Welti/Harman papers assuming identical signals and subwoofers at each location. No delay, level or EQ differences were applied. I have measured more than a few rooms where the recommended Harman approaches resulted in huge nulls in the listening area. The space has to align with the assumptions made in the study, and many do not. Note this system has an F18+ & Slave at the front wall, and another F18+ and Slave at the rear wall. Fortunately adding delay as a range of flexibility for the front/rear grouping of subs is often enough to achieve very positive results. This assumes that the front and rear locations are reasonably complimentary and useful, which is most often the case for front wall and rear wall placements. You could very likely slide either pair of subwoofers to the 1/4 & 3/4 width points with modest changes in the results. A large delay is not always needed on the closer subs, and sometimes no delay is needed at all.
Seaton Sound, Inc. firstname.lastname@example.org 773-290-8436