What if instead of doing a traditional screencast or traditional howto, you create a hybrid. A how-to document with a short video explaining each section of text. That way steps that are trivial can just be read through, and where a demonstration is needed, the user can watch the video.

The problem is, of course, that few people will find the same steps trivial, so you'll have to do double the work (video and text). On the other hand, this will solve the problem of pacing.