Update findThreshold docs for gmm/dens trade-offs

[Steven Kleinstein]

Get the per sample distToNearest plots from Jiawei... Looks like dens performs better for large sample sizes and gmm fits better for low sample sizes.

We should update the docs to explain the trade-off, ideally with an approximate number of sequences under which performs optimally.

[nima nouri]

Resolved by commit: ba9ae72