Yes self-discharge is slower at 3.6V than at 3.85V, but it's already very slow at 3.85V, at least compared to 4.20V.
You don't need to leave margin "in order not to self-discharge into death", that's a myth which only applies to cells that are so bad they are going to die anyway. I have tested storing cells at 0% for 1.5 years, the voltage after storage has bounced back to higher than it was when starting the storage. They just don't self-discharge at all. I still won't recommend storing at 0% just because that is not usually done and there might be some reason I'm not aware about; just sayin' in my limited testing I haven't seen any ill effects storing at 0%.
Buying cells to build pack, I have noticed that Samsung had changed their factory charge level (at which cells are shipped) from 50% to about 30%, somewhere in 2018 or so. I don't know if this is to increase safety during shipping, or to increase the shelf life (which at 50% was already great), or maybe there is no special reason to do so. Just saying that the standard shipping charge level is between 30%-50%.
Anything between 3.5V to 3.9V is quite acceptable for long-term storage but between 3.5V and 3.7V would be even better.
Self-discharge in itself isn't problematic, irreversible damage is. But the two go hand-in-hand, in high state-of-charge and/or temperatures you have both elevated self-discharge and elevated capacity loss rate.
Below about 60% SoC at room temperatures or lower the self-discharge in good quality cells can be so low it's completely negligible and you can say they have no self-discharge. In my measurements, all measurements in such conditions were lost in noise, i.e., I would have needed way longer period of time than just 1.5 years, maybe 10 years to see how fast they actually lose charge.