There are often reports about breakthroughs in alternative technologies and when I look, I can find charts claiming that in 2014/2015 top solar conversion efficiency is reported to be at about 45%. However, when I look harder I don't seem to find obtainable Panels above 21%. The highest grade I could get were the Mobile Technology MT-ST110 panels on the odyssey. I can't help but to ask myself:
Now, a couple of researchers with interdisciplinary and seemingly non-related backgrounds published the concept to apply a quasi-random nano-structure to a solar (PV) cell in order to increase the absorption of photons (decrease reflection) and reported an overall broadband absorption enhancement of that solar cell to be 21.8%, when a Blu-Ray land/pit pattern is applied.
From the little info publicly available, it seems they've just tried different materials for this approach and discovered that Blu-Ray's compression algorithm (tested with Jackie Chan - Police Story 3: Supercop) creates a favorable quasi-random array of lands and pits (0s and 1s) with feature sizes between 150 and 525 nm, which seems to work quite well for light-trapping (photon management) applications over the entire solar spectrum.
http://www.mccormick.northwestern.edu/news/articles/2014/11/ http://www.extremetech.com/extreme/194938
IMHO, it should be easy to use a real random number generator as a source, create some land/pit patterns following Blu-Ray specs (physical), run some tests and apply that to the production process. If it works, the sudden increase of convertible sustainable energy would be more than just considerable.
Let's come back here in the future and measure how long it took to get into a real product. Especially in constrained off-grid systems like Apollo-NG, with such little usable surface area, every 1% of efficiency counts for a lot.