I do all my editing in PD, I just don't like PD to do much encoding, I'd rather it use SVRT when possible. My work flow is as follows and should address your question in detail:
1) I take 100's of raw Canon HD mts files (via the playlist) into a utility to mux them together a create 1 large file, call the file raw.mts. Takes only a couple of minutes as no rendering is done, just file joining. This alone can speed up PD a fair amount as PD does not like lots of small clips in the TL for editing, during Produce, or Create Disc. Even PD10 with the new smarter SVRT with many files during create disc discussed in a bullet here
http://forum.cyberlink.com/forum/posts/list/19428.page#101990
2) I take file raw.mts created in (1) into a utility to render a MPEG2 file which will be compliant with PD SVRT DVD HQ settings, file render.mpg. I do this for video quality as I've not been happy with what PD does in converting Canon HD 24Mbps raw to DVD HQ. Too many artifacts, pixilation, jitters, shutters, quivers for me....these all depend on what you film though so many users might be happy.
3) I use file render.mpg created in (2) inside PD timeline to do all my editing
4) I create a DVD folder with DVD HQ, SVRT is used behind the scene so my render.mpg is virtually untouched except for edit areas that need to be encoded. My 30-60 minute DVD HQ folder often created in under 5 minutes.
5) I now have my good DVD HQ master Folder for my video project.
For the HD disc I:
6) Manually edit the pds file from above and change reference pointer in the file of render.mpg to raw.mts. I now have a new pds file with the identical edits and identical footage, just all HD now.
7) I create Blu-ray folder with H.264 (raw footage spec) and now I have my Blu-ray master Folder. Again since I used the same create Blu-ray folder video spec as my raw.mts file, SVRT is used and folder is created in minutes.
The above workflow offers the following advantages for me:
1) I'm always editing DVD HQ quality mpg footage so editing CPU load in PD is significantly reduced. My system just flys with this.
2) I edit only once and end up with the capability for both a DVD HQ and a Blu-ray H264 disc that PD does not encode (SVRT) for a large portion of my video. Only edit areas are encoded by PD.
3) I keep all my stuff in various folders all with the raw video files processed as above. So if I pull in some footage from another folder into my current pds project, that folder structure will also have the HD version there as well. So when I save my final pds file and edit to convert to HD, all my folder structures have both the DVD HQ mpg quality and HD mts files so no issues.
4) I only absorb the time to convert footage once, when I combine multiple footage clips nothing ever needs to be encoded again, it's all SVRT compatible to make either a DVD HQ or Bluray project.
The workflow above works for what I need to do, everyone’s mileage may vary. I produce about 60 unique discs and about 900 DVD HQ copies and 120 Blu-rays copies a year and can not be messing with DVD quality or work flow issues on 60 projects, I don't have enough hair left!
Jeff