I read somewhere, years ago, that species of birds that live in flocks tend to display better group-memory. I don't remember (pun intended) how that was measured, but have observed this to be also true of human interactions. Aside from mere cognitive stimulation---in a group we also tend to interact more, so there's a greater chance of certain facts being continually refreshed---I believe that there's more to what enhances group memory. That is, I think that there's another factor that influences group-memory, not by virtue of cognitive stimulation alone, but the kind of stimulation that has greater likelihood of occurring as the group size increases.
It has to do with both the abilities and limitations of each individual, and the fact (conjecture) that those abilities/limitations vary significantly among individuals. The hypothesis underlining this idea is simple---it's very plausible that given that we've been a social species for most of our evolutionary history, it would appear to be an adaptive trait to distribute cognitive abilities related to memory and recollection over the entire group. In particular, there may be a significant variation of the temporal ranges that individuals are capable of recollecting proficiently. That is some people would be better are remembering recent events, with high fidelity, whereas other's mid-term or long-term memory may be much better. After all it's the sum of all memories, fed into the social collective that counts, so it needn't be preserved by each individual---that would be an unnecessary over investment of neural-power.
From careful observation of my family and friends I have noticed that some of us display great proficiency in recollecting great amount of detailed information of current affairs, i.e. information gathered in the recent days or weeks, but may struggle holding that detail for a long period of time, and their memory of current affairs progressively fades over time. Others, on the other hand, display much better long term memory than others.
Perhaps such traits can be shown to be more generally diversified with respect to temporal ranges of optimal recollection. Moreover, there may be variance regarding the kind of of information that is remembered---some people may be better at remembering how to perform certain actions, others at phenomena that occurs in the surrounding environment, and others still may be really good at keeping a moral tally of group members, i.e. those displaying altruistic actions and those that exploit the benefits offered by the group. Combining those individual abilities in a social group would obviously enhance the group memory, simply by virtue of mutual compensation of individual limitations. So, such a social mnemonic mechanism would result in the group super-organism as a whole displaying much better memory than each of its members alone, on the condition that information flow is unhindered.
Those ruminations on the nature of memory are an unintended interest of mine. I just can't help but make conjectures about it. I've explored this subject matter before, via a short story---in particular how perception influences our memory and recollection.