I've never been much of a rewriter.
I'm usually the sort of writer who obsesses over each word before it hits the the page. Thus I I rewrite less than other whose first drafts might have been a little breezier in the making. For me, completing the first draft represents the bulk of the work. I'm thinking about a lot of stuff during its creation that a lot of writers leave for the latter part of the process.
I never planned it that way. It's just the way my writing habits evolved. Polish the first draft as much as possible - fix problems that arise along the way, including going back and rewriting chapters as needed to resolve continuity issues and the like. It takes a lot longer to get a finished story, but then you only need to iterate over the writing to bring the rough parts up to the level of the good parts. Major problems should be mostly taken care of by that point.
With Clanless I'm trying something else, as I believe I've mentioned in the past. I'm attempting to focus on getting the story onto the page faster and trusting in my ability to rewrite it up to quality. I don't know if this method will be faster in the end (and to be honest, the patchy amounts of time I'm able to devote to writing make it harder to judge) but I do want to give it a try.
So far, progress has been... odd. I've skipped entire element of the story to fill in later. I have a myriad of bolded passages telling myself what needs to be added, or what will need to be reconciled if I end up going a certain direction later. I also have a number of passages emphatically marked for replacement because even as I finished them I was cringing at how bad they were. Normally, that kind of thing wouldn't stand with me, even in a first draft. All in all, I estimate I've left a good 15k-20k words to be "filled in later." That's out of 35k words written so far.
Yeah. Not my normal MO. I guess we'll have to wait for the final product to see how it the process helps or hinders - and even then, it's rather hard to control for the effect of any single variable in the writing process. Just too many to account for and "life" is not the most stable experimental environment.