I keep an open mind, also it's a well known fact that solar cells performance degrade with increase temperature so it doesn't take a stretch of imagination to see if u can control the temperature and absorb enough reflective energy you can beat a straight facing non-tracking solar array. Notice his "tree" is next to a white reflective wall which allows massive amounts of bounced light, either way I think there is something there.

Look at his design it isn't tightly packed, it colluminar loose packing much like a tree or plant. Put that next to a highly reflective surface and you might indeed beat a straight facing solar array esp since the straight facing one isn't tracking so it will only optimally absorb within a narrow time window vs the "tree" which will have much longer time of absorption enough though be less efficient. Again that's why trees have leaves all around in hemispherical because they don't have tracking. Why don't trees and plants have leaves just one level deep? Because as Yann L pointed out they can't absorb solar energy above a certain lum , the light diffuses through the first layers and they capture the indirect bounced and transmitted light. Same here, multiple level redundant arrays arranged in hemispherical random way will probably beat a straight facing non-tracking array as long as there is enough bounced light.

That's my theory anyways.

-ddn

It really depends on the plant/tree. Many trees have leaves all over because they also serve as structural stabilization transfering what would otherwise be strong vibrations that would cause the tree to fall over. If you look at smaller plants many of them have leaves that face generally the same direction and will even track with the sun through the course of a day.

If you look at the blogs that go against it, they show why having multiple solar panels all pointing different directions couldn't be as efficient regardless of the situation. The gist using your example of the wall (if pointing them at the wall were truly optimal) being that pointing them all at the wall spaced out enough to be cool would be more efficient than putting them all on a tree with only some pointing at the wall and some pointing randomly in other directions; rather if every individual node is performing optimally, than the whole should also be performing optimally. There could be a lot of leeway for situations where nodes might not be performing optimally, but I feel like a tree like setup still wouldn't be the best way to go about it.

That's my theory anyways.

Ok, I understand your position in this specific example. Nonetheless, there are a few issues for this theory to work. I want to point them out as people is clearly not proficient with photovoltaics.

Crystalline solar cells suffer strong performance degradation with heat. This is true. Does not apply to amorphous, nor CIS (well, it does, but much, much less).

The amount of heat we're talking about is so massive that convection, no matter how good, simply cannot cut it (for standard panel designs). The amount of airflow required is so massive the cost and consumption of the fans are more or less on the same magnitude of the gains. Whatever the net balance is positive needs to be proven. Next year, I will experiment with water spraying at regular intervals since I have access to low-cost "crude" water.

In terms of solar reflection, standard "white" is basically not so good. Companies also sell solar-grade white (as well as solar-grade black). I don't know if the document I've seen is public, but a company gave me a report stating the benefit of solar-grade white was about +15% when using Solyndra's panels when compared to their standard white. That took me by surprise.

While the cell itself might eventually be similar in transparency to a leaf (never seen one in open sunlight), typical panels are not. Hybrid panels are basically opaque.