Overlap, and any modularity/clustering you build on top of that, does not capture if species are very close together but do not overlap.
How much does this matter?
This might be exactly what we want, esp. in cases where the non-overlap holds up to upsampling.
I'm somewhat concerned re: the scaling of sd with mean. Larger species are going to have a harder time being distinct from each other (but again, this is perhaps what we want?)?
Overlap, and any modularity/clustering you build on top of that, does not capture if species are very close together but do not overlap.
How much does this matter?
This might be exactly what we want, esp. in cases where the non-overlap holds up to upsampling.
I'm somewhat concerned re: the scaling of sd with mean. Larger species are going to have a harder time being distinct from each other (but again, this is perhaps what we want?)?