Any evenly spaced LES grid in the size areas hobbyists generally use will have a hot spot. You can minimize (to a degree) the difference by using more light sources at lower powers and closer canopy distances (strips, boards), but there will still be some drop off at the edge and reflective loss goes up. Still, this is a useful way for the hobbyist to get more radiant energy in a given space while keeping their hot spot to a minimum.
Even with the low wattage higher density cob arrays I've built for myself, I prefer to run them at 12". At 6" from the wall and 12" above canopy I'm getting 80% at edge and 120% at center. This represents some reflective loss over the intended hanging height of 6" but after ditching the panda plastic for mirrors and mylar I feel pretty good about it. I get readings between 800 and 1200. So I don't think there's one style that's the best, but the best height for a lamp is the one it was designed for. I always designed my retail lamps to hang at minimum of 12" and max average ppfd of 1250.
If I want to push the limits and hit the holy 1500 PPFD, I don't want to see an 1800 PPFD hot spot in the center. If I design a 1250 PPFD light array I will expect the hotspot to be 1500 and the edges to be 1000. If I was okay with edges at 800 and 1200 in the center I would plan a 1000 PPFD setup.
A more extreme solution would be to make the light grid asymmetrical or use 60 degree or even more narrow optics to reduce center intensity and accent the sides and especially the corners. The most logical reason for considering such a strategy would be wanting 1500 PPFD complementing CO2, being less concerned about electrical usage and more concerned about how far you can push the plants in a given area. This would ideally involve design software or a modular rig for testing with a light meter prior to making the design permanent. If you begin to consider the 3d aspects of it it becomes even more complex.
