Ok, seems like at least that "concept" is mathematically/theorectically proven.
That the diffused reflection is safe, but if and only if when relying on blink reflex (0.1s).
I feel we also need to answer the following very important questions, which actually i and most have mentioned earlier anyway :
(1) When we are shining the laser around on the mentioned non-mirrored/safe surfaces, are we somewhat overriding the blink reflex? (0.1s). It seems that the photochemical reactions on the retina is accumulative. Let me raise another example, in the astronomy circle, some lay people suffer from localised blindness (during solar eclipses even without looking through the telescope/binos, and they don't the eclipse shades so just use their hands to shield and squint their eyes).
I am sure quite a few like to focus their lasers indoors to do popping, squint at the dot etc. For eg, is it safe to try to focus the dot at 3m for 20 seconds? You'll have to multiply by 200x (20 seconds over 0.1s blink reflex), something like that.
Most likely we'd need the professionals to see if we missed out anything important. They'd far more "stories" to share. Its the same in your own respective professional fields, in which reality deviates from theory. Aka Murphy's law.
My apologies if i seem skeptical, because in reality I already experienced some loopholes + close calls....simply because in reality it is never a controlled environment as our usage is really for hobby use.
Heck I already had some close calls with my 100mW Nexus a couple of years ago.