Once you submit your paper it is visible to experts you probably know (or at least know of). The double blind process does not work when the subject matter, and writing style, and citations correspond closely to specific labs or researchers. You stake your reputation as well as that of your lab/company from the moment you submit.

I say this from personal experience, I once published last minute for a conference and was violently chewed out by my bosses because of things they disagreed with in the paper.

> The paper's only in review

You submit a final version for review. You cannot change a paper significantly after it was accepted.

Conferences of this kind would not allow major revisions unless they were requested by the conference's reviewers.