N Butler,
What length of timber are you needing for your plates? I don't think there is anything wrong with a good scarf joint when needed.

Jay,
You are probably on to something with the layout and execution having a lot to do with the buildings coming together straight without additional help. The arrangement of plate above posts with rafters joined to those plates still seems preferable to me than interrupting the cross members at the plate. I could not put my finger on a quantifiable reason at first but I after some thought I think I have one to field.

Lets consider this joint intersection under racking pressure. We will also assume that there is knee brace from the post to the rafter carrying horizontal member and there is a rafter placed directly over the post.
With the horizontal member broken up by the posts the only thing resisting the racking pressure (including any levering action imposed by the brace) is the joint itself, which by definition is a potential pivot point. I'm however not implying that the joint can't be designed to properly resist this force.
With the plate passing over the post it gains two critical advantages. Gravity, (always a force to worked WITH when possible) is on the joint's side, augmented by whatever load is transmitted through the rafter. Also, the string of post to plate joints is going to working more in unison to resist the destructive pressure. If interrupted by the posts, it would seem that they are working concurrently but not really "together" if you catch my meaning.
Any scarf joints necessary to the arrangement would have to treated as a separate variable entirely, as they are all over the place structurally in how well they mimic an un-scarfed beam. But I think that we can agree that it is possible to produce scarf joints that perform highly enough to not detract form my argument.
This is just my own thought experiment based on my own notion of structural "rightness". It just seems more "right" in my mind that the plate should pass over the posts, and I don't even think for purely "historical" reasons. I imagine that an engineer could point out some fallacies in my thinking, which I would love to hear! Has any comparative study been done that anyone knows of? Could be an interesting joint busting demo at the net conference!....