Ah! Well that's a surprise. Cinema seems to have a magic checkbox for just about everything - I assumed something so basic as fixing sloppy modeling techniques would have been resolved long ago 
When I started this project I was working fast and dirty - assuming all could be fixed in the print. Ha! Lesson learned.
Ha! Is there a command for removing the overlapping polygons and patching the gaps? The optimize command doesn't appear to do anything when I apply it. When I try Weld, everything goes nuts. Shooting in the dark.
Yes, I considered using the volume builder, but I haven't had good luck with it for geometric objects in the past. However, your sample looks pretty sharp, so I will give it a study. It would greatly simplify the prep for printing on the modeling side (though it does take a while to process).
What does it mean for an object to have "open edges?" I optimized the extruded interior block but it didn't seem to make a difference in how it was handled. Unfortunately, mesh analysis has not been a strength of mine.
I sent a model of a building to ShapeWays for printing, and it failed the mesh check because of open edges (see attached). The yellow lines identify the offenders. The attached model includes a window and an interior block created to beef up the wall thickness for printing - both of which show up in yellow on the analysis. The interior block was an extruded polygon selection. And I suspect the windows were created with a similar process, then cloned for placement.
I did a mesh check but didn't get anything I recognized as an alert. So I'm not sure how I would go about finding/repairing this sort of thing in the future.
Thanks
Back again with the tripartite arm system...
I want to use a Loft to sheath the splines of the articulated spline arms like in the example, but the Null in the construction apparently breaks the effect. I thought maybe a Connect object would help resolve it, but no such luck.
I don't want to resort to point animation with pose morph because I want to drive the animation with fields, but maybe that's the only option?
Current project here:
I keyframed the cloner into submission and then moved on to the next step - extruding the splines into shapes. However, applying the Extrude to the spline offset the keyframe positions. I've tried every trick I know to get the extruded splines back into alignment (without re-keyframing) but no avail. I tried the constraint tag, but that didn't carry over to the cloner.
It makes sense that with the Extrude as the parent, it would break the dependencies, but I thought using option-click to add the Extrude would transfer all the axis information to the Extrude, but it doesn't.
I used the polygon pen to create a new polygon on top of the spline, and that stays in position. But I'd prefer to keep it editable as long as possible.
Current version:
Arm Builder_v5 Idealized_v3.c4d
Why does the cloner track the end goal of your original rig? My experience is that cloners do their own thing regardless of what you want them to do.
Back again 
There's a mechanical aspect to this project that I hoped the dynamics would resolve, but that seems more difficult than I expected (still planning to mind meld with the tank tread tutorial). In the interim I wanted to get a better sense of the mechanistic aspects of the moving arms, and so adopted a much simpler, "brute force" approach (i.e. keyframes). This is what I got.
The stumbling block this round is the enigmatic "Start_0/Move_Me!" null with the pose morph tag that you provided in the last round. It works perfectly - however I can't for the life of me figure out how or why. On my new version, the cloned arms do not track with the end goal (which isn't a goal this time since I dispensed with the IK chain in the "brute force" approach) so the "Result120" Xpresso rig is broken. Leaving the Cloner to follow the keyframed straight-line trajectory (as expected) which doesn't track with the adjacent arm. In other words, I broke it
I think if I reverted to a IK connection between Arms B and C as before, and used the Arm_C Marker as the goal, I could probably get the pose morph to work again. Or not? The angle for Arm_B seems then would be the driver for opening the configuration.
I have a dim appreciation for how I could resolve this with pose morph, but I'm just not getting it yet. Sorry!
Fantastic! Yes, the tank treads are exactly right. I was looking at a Rocket Lasso tutorial on chains, thinking that might offer some options. This is much more directly applicable. Thank you!
Yes - feedback loop! That's the big problem. Would it be better to use connectors and springs instead of joints? Not sold on joints, just too old school to know what else might work.
