I have to admit though, out of selfishness, I do not use that measurement configuration any longer.
Since I'm selfish, FR at the MLP is my priority, and I found out that I can get a good FR at the MLP, AND save myself time with a much simpler measurement method (which of course, isn't recommended by Audyssey).
I simply place the mic for position 1 at the MLP on the couch, dead center between the front L&R and ~30" from my back wall. I use a low profile boom mic placed on the cloth couch. Position 2 is measured 16" forward of 1. Position 3 is the same as position 1. Done.
The results measured at the MLP
P.S. in contrast to kgveteran's graphs, Graph 2 in this post actually shows real nulls at ~150Hz and ~185Hz. In order to show a null, you need to see the graph with NO smoothing applied. It will be sharp, the dip will be deep, and it will center over a frequency. EQ will not correct it. The only way to correct a null as I mentioned in an earlier post is by a) adding bass traps to absorb and reduce the amplitude of reflected bass waves, thereby reducing the intensity of their destructive interference, b) repositioning the sub to alter the cancellation of the reflected and direct waves, c) adding subs to excite more/different room modes