does communication between api, grpc and ui cause a hairpin scenario?

[wibed]

coming from a helm perspective the minimal requirement is to pass on a baseurl, which the components use to communicate
with eachother.

doesnt this cause a hairpin scenario?
is there a reason not to configure the coredns instance running on the cluster, or even setting one up?