The "best" position is lots of them. To get an idea of what is happening around the various seating positions you have to measure in them.
Now what you can do about it, depends on the particular system you have.
In some cases you take a bunch of measurements and then just a
"average" them or determine what seats are more important than others.
Regarding room modes, a bull dozer is usually the best tool for the job.
No seriously, I'm not kidding. And I have told more than one facility that the best thing they can do is to tear down teh building and build a proper one for what they are tying to do in there. They don't like that.
Room modes are another reason for multiple mic positions. To see if what you are reading in one position is the same all around. If it is then you can do something about it. but if-for example-you see a peak (or a dip) at some positions and the opposite at others, there is generally nothing you can do-unless those are caused by interactions of the loudspeakers themselves.
And then you can only do something about it-if you have the control (DSP-different amp channels etc) and the knowledge to realize what may be causing that dip in the response-overlap-time offsets-phase errors or the loudspeaker itself.
The first thing to do is to get the system aligned as good as possible-paying particular attention to the phase response.
Then you can do a feedback tuning on the fixed position.
The roaming mics are a totally different animal. See if you have the same feedback freq in different places in the room. If you do, then use the console eq to notch that out-NOT the house eq.
But get the system as good as you can, BEFORE using eq to try and get more gain before feedback. Don't jsut start whacking away at the eq