Brian(Bigdog), one of the contributing problems in this area is terminology. There's no "7.2" currently in existence(despite manufacturer labeling of receivers with two sub outputs); there's only one .1 LFE(low frequency effects)channel and using multiple subs in the channel can't change that. Using two center channel speakers in a 7.1 system doesn't make it 8.1 and two subs don't make it 7.2.

This one channel can't be calibrated independently for level and distance delay for two subs by most receivers(including your SC-27). If the subs are the same model and at the same distance from the listener, the one set of adjustments done by the auto-calibration should be entirely satisfactory. If not, the levels of the subs should be matched manually(not by ear but with the aid of an SPL meter)before the auto-calibration. The auto-cal will then correctly match their levels with the main speakers. The distance delay set by the auto-cal won't be accurate if the sub distances from the listener are significantly different, and probably the best that can be done is manually adding the extra distance to the farther sub after the auto-calibration.

There are receivers such as higher cost units using Audyssey Sub EQ HT which send separate level and distance delay settings to the two subs connected through the separate outputs and then after the calibration make room EQ corrections to the composite output of the two. Dr. Kyriakakis has discussed this among other factors in his excellent thread here . Among the findings were that separate room EQ adjustments to the two subs weren't as good as a uniform set of adjustments to both.

As to the location of the multiple subs, the best location for each individually isn't the point, but rather locations where they have opposing acoustic parameters such as to reduce the effects of room modes. The research of Dr. Toole and his associates at Harman indicated that mid-points of opposite side walls or the front and back wall, or diagonally opposite corners gave good results.

Enjoy the music, not the equipment.