My problem with random placement is that it is, well, random. I do not see how you can assert with any confidence that a random placement is better than an ordered placement. You can say that a random group arrangement will smooth the sound field by averaging, but that can be said of an ordered array too. In general, the more sources the better, whether random or ordered. But the question remains, where are the best places to put the subs. I am not convinced that a random arrangement is best.

What I'd like to see are 3D energy distribution plots of each of a handful of test setups. Not just the average and deviation, and not just the plot of a single position or small area in the center of the room. I'd like to see energy distribution charts of the whole room, showing various room sizes and speaker placement configurations. So far, Welti has provided the most data. Perhaps his study is incomplete, but that has yet to be seen.

No, the right thing to do is to send a mono LF signal, and then decorrelate this signal between several sources on playback. That way there is no degradation in the case of a single sub.

I did read enough to know that you should be carefull in interpretation of these results. Thats because they generated ideal signals that were correlated and uncorrelated and we cannot do that with real sources like CDs.

There was no discussion of "preference" only "audibility". So it cannot be concluded, per se, that they would recommend front to back or side to side placement, only that there is a possibility that they will sound different.