I don’t know the insights, but i think it would use more memory.
when you load a sample in the multisampler, it can’t know if it’s the same sample you’re loading, or a different one.
so in my understanding it would load the sample again.
i haven’t tried (gonna check that out) but i assume that when you store the multisampler preset, inside the preset folder it will create all the files with the same sound:
i hope so. not saying that’s not possible. they might be calculating a hash for each file in the preset and effectively recognize the sample is already loaded in the pool.
they might have thought to optimize memory in some similar way.
haven’t had the chance to test it, i will when i have a few mins.