If using qemu i440fx machine types, the hda sound model is ich6. If using q35, the sound model is ich9. However to end users the term 'hda' is far more well known. We should accept hda on the virt-install cli, and expose it in the UI, and just map it behind the scenes to the appropriate sound model in the XML
We do expose it in the UI nowadays. The cli mapping isn't really that interesting IMO, users should just take the default and virt-install should aim to give the best model, or users can specify '--sound default'