Bdgr is right. I had a teacher in college who was a tech for the company who supplied the cluster they used for keyboards. The first day he worked on the show, he asked the engineer how the line array stuff worked. When he was told, he didn't believe it. The engineer gave him a dB meter, and told him to walk away from the stage while some music was playing. My teacher started at the front edge of the stage, and walked away, watching the meter. The volume would fluctuate by less than 3 dB, total, all the way to the back of the arena. The problems with the “Wall of Sound” were the balcony got almost no low end, and it was ridiculously expensive and time consuming to set-up and move. They stopped using the "wall" the first day of the energy crisis.
There monitor system was the front of house system. Every thing was right behind them, so they heard everything through the FOH speakers. The mics were wired with reversed polarity so they would cancel anything which hit both mics. You pick one mic (it does not matter which one) and eat the mic as you sing. It works great. If you don't believe it, try it. Take a couple of 57s and hold them next to one another, sing into one, and eat the mic. By the time you are 1 inch away from the mic, you voice is canceled out, but if your lips are touching the mic, it comes through clearly.
One last comment on line arrays. There are a lot of different speaker manufactures coming out with line arrays these days. Even Meyer is making a line array now, and for a few years they were trying to fend off the current trend towards line arrays for event PA systems. The problem is this. They are NOT line arrays. When the V-DOSC system came out, it was designed and spec'ed to be hung in a straight line, perfectly vertical. This is a line array. But pretty much the first time anyone tried to use it, they found that they could not hang the arrays straight and get sufficient coverage of the front rows. The fix they found was to arc the bottom of the array. The only thing is, as soon as you arc the array, the coupling which was taking place when it was vertical ceases to take place, and you start to get some serious comb filtering. The other problem with line arrays is that they, by there very nature, narrow the width of the dispersion from the cabinets. In an array, a cabinet which has a 90 degree dispersion is reduced to about 70 degrees, or there abouts. On top of all of that, I (personally) find the modern line arrays to sound really inconsistent and flat. Give me clusters, any day.
Light
"Cowards can never be moral."
M.K. Gandhi