If you are going to edit in the DV codec there is no benefit in options 2 or 3. Just output to mini DV and have it dubbed via SDI to Digibeta at a dub house.
There is no quality difference in using DVCAM.
However, if you edit in the DV codec, but copy your final edit into an uncompressed sequence (you will have to render everything), you will see some benefit in composited graphics provided you master to Digibeta via SDI. So option 2 would work in this instance. However, you would need drives capable of playing uncompressed video.
If you can online on an Avid or FCP system uncompressed, you might see some improvement in graphics compositing, but only if you do not edit in the DV Codec.