The word for when corals get too much light and start to grow less/zoanthale produces less food, is photoinhibition. Essentially the zoanthalle produces more and more food with more and more light to a point, but at that point they start to slow down, and farther past that point production will actually decrease. (Calfo talks about this in his propagation book)
Second on what Jay said, He showed MH putting out much more light, but also that the levels produced by T5's tested were perfectly adequate for all sorts of corals including numerous SPS.
The thing I found particularly interesting was the thing about light reflection (and this was why the PAINTED backgrounds are bad). He said the super simplified explanation was something to the extent that when ever two clear but different things meet there is reflection. ie water touching glass reflects light, and glass touching air also causes reflection.
Because of this, he found that when shining a light into a fish tank full of water, the point where the water meets the glass reflects light back into the tank, and the other side of the glass where it meets the air also reflects light back in.
Because of the above;
-painting the back glass will eliminate the reflection effect coming off of the outside of the back glass (enough that it showed dramatically in his light distribution graphs). He also stated that as long as the background isn't directly touching (painted on, or stuck on with adhesive) it DOESN'T cause the loss of reflection.
-This glass reflection effect also favors T5 over MH as the distribution of light from a point source doesn't bounce as much light off the glass and back into the tank.
-Letting the glass get dirty actually will decrease light intensity in the tank (so go scrape all that coralline...)
If you want a background, don't paint it or use stick on backing, BUT you can hang a curtain behind the tank, or any other option that still allows there to be air against the glass.