I believe it can. Data is chucked, so different data may share the same chunks.
It is more about where the data is stored. On safe network, the chunk hashes dictate which section they will belong to on the network. The nodes associated with that section then take ownership of the data and supply it on request.
On IPFS, as far as I gather this is completely different. The hash of the chunk doesn’t dictate where it will be stored. Data exists initially on the node it is uploaded to. It may be pinned by others on an adhoc way. Caching is then used to push popular data around the network.
So, on safe network, if the cache doesn’t contain the chunk, it is easy to find the section responsible for it and request it. On ipfs, I gather nodes have to be asked if they have a copy and a network search commences until it is found (or not found, as no perpetual guarantee). I was trying to dig deeper into this the other week, but this is as far as I got. It sounded pretty rudimentary compared to how safe network handles this.