Mic 2 ->left Mic 4->Right

If you're in a hurry, skip the even number tests. I do the L/R swap
in order to detect RF and other subtle stuff that show up. Sometimes
I have to do the test over when they do.

 From each of these "takes," extract the quietest moment. Place these
moments in a mixing time line to be able to play all possible
combinations.

By having the mics close together, they should create a dominate mono
image in the middle. Anything you detect on the far sides is unique
to that mic.  Hiss is the easiest to hear/distinguish. Differences in
the low-end can be detected but not if you have fairly steady local
traffic.

After you have narrowed it down to the best pair and the worst pair.
I put the mic pairs in my favorite stereo rig and test: