A lot of different ideas expressed here.
To clarify, I'm not talking about self-discharge, rather overnight discharge due to parasitic loads. Mine is relatively small at .3 amps. The Trimetric shows an overnight loss of ~4 Ah. Golden HVAC regularly posts about his 35 Ah per day loss to parasitic loads. Assuming that his battery bank is about the size as mine with half of the loss occurring when solar is off, that equates to a 4% loss over night. Not sure why some think this is such an impossible scenario.
As to this merely being removal of the surface charge, that may be. I certainly don't claim to be a battery expert (I would be giving advice instead of asking here) but surface charge to my understanding is a localized phenomena due to the chemical reactions of charging not having time to balance throughout the plates. I still believe that you can't discharge a battery even slightly without a chemical reaction taking place.
If this is indeed just another shallow cycle then nothing to worry about. A typical depth of cycle vs. available cycles indicates the shallower the better. However it doesn't show what happens below 10%. Another glance will also show nothing magical happens at 50% either.
