That sounds good. As the general purpose of sectioning is to created detail views on layouts, I suppose this should automatically create a named view that can be called from a detail.
That is correct and it is the feature request that I've voiced in the past as well. Objects and layers should have a property / collection of properties to set: X-section line type, X-section line width, X-section line color, X-section solid fill, and X-section hatch. All objects that reside on a particular layer (and that do not have a By_Object or By_Parent or so override) will automatically get these properties applied when a Clipping Plane intersects them. I suppose that for this purpose we can recycle the "Line Type", "Print Color", and "Print Width" properties that already exist, unless, of course, a user has curve entities on the same layer that he'd rather not have behaving like a cross section.
Again, as the purpose is to create details on layouts, would it be an idea to automatically place the dynamic 2D layout on a layout instead of in the modeling environment? It would then live on the layout itself, not in a detail.
See above [X-section properties]. It's been a while since I've used SectionTools but I suppose that the cross section curves that are seen in your images are actual curve geometry. I have for the longest been hoping that Rhino didn't have to create actual geometry, that X-Sectioning could be some display trick or something. We see that creating the actual geometry leads to a substantial slow-down when compared to detailing in parametric software where the details are obviously 'just' taken off the render mesh and not the underlying model. Also, in Rhino, the Clipping Plane gives us what we want (*) on the fly. So to the uneducated like myself it appears that pretty much is in place.
[(*) except for snapping and that often solids are not filled]
At any rate, that has been voiced before so I guess that it has been looked into and deemed not possible.
Still, I would much rather not have to deal with the extra layers and geometry that the cross sectioning functionality introduces [layer management isn't that much fun in Rhino]. Could it be that this is kept entirely behind the scene - except for the fact that one can see it? So that way, the section curves would get their properties from the X-section properties that are set on the layer where the object resides (or object itself). One would be able to snap to the section curves (also a request that should be fairly high on the ladder by now), but not select or move or anything. If a user at some point and for some reason would like to do something else than to just illustrate / annotate / detail a 3D model, there could be an ExtractXSectionGeometry command that rips the curves out from behind the scene and places them into the 3D world.