One DNA sample isn’t enough. You need dozens or more to repopulate a species. If you just had one or two, you would have to rely on inbreeding to repopulate which would create too many bad genetic problems from bad genes.
Whether this means the bottleneck hypothesis is overstated, or that we're looking at a tremendous number of bottlenecks, isn't clear.
Vs large mammals where raising one genetically-hobbled offspring (who then dies, or fails to reproduce) is a huge loss to the potential of the species.