Multiple channels per sound vs sound per channel

Is there any difference in behavior between having a channel per sound vs a single sound and multiple channels when trying to play multiple instances of a sound at potentially different positions?

Currently I am using the former but it makes thing more complicated, but concerned that latter means that multiple channels could cause lots of seeking or some other unexpected behaviour.

This post answered my question