Thanks for the responses!
I have been experimenting with using both the complex and flatten publish methods to export to resolve for color grading.
I am able to get the footage an re-assemble using the edl into resolve from smoke using those methods, where we could then do the grading. However, if we would want to then round-trip back to smoke, that could be a little more tricky.
Smoke seems to have some issues relinking to the new media from resolve. I believe in most cases we wouldn't need to go back to smoke once we have graded in resolve, however I could see a few reasons why you might want to before the final output.
We would be using resolve as either the intermediate step or end step. We find that in our pipeline most clients want to sit in on the color session, which would mean they want to grade the final spot, not just raw footage before it has been conformed. So in our workflow, we would want to be able to conform all the footage into the final spot using Smoke, that way any timewarps, split screens, fx, etc. would be handled using Smoke. Then smoke outputs the flattened/complex footage using the publish feature. We would then theoretically reassemble using an edl in resolve and do our grading.
The reason for this being that the conforming features seem to be pretty limited in resolve which is why we would take advantage of smoke's tools for this part.