I see your point, but I respectfully disagree. Perhaps is is just a difference in perception on the function of the limit mate. You see the limit mate as a mate that requires motion in a specified range. I see the limit mate as not requiring motion, but rather to limit how that component can be constrained in an assembly.
Getting back to the cylinder example, I think limit mates should allow one to apply the limit mate, documenting the cylinder's limits. Then, other constraints (weather they are in the cylinder assy, or up a level when used as a flexible subassembly) should be allowed to fully define the cylinder length. Thus, when the length constraining mate constrains the cylinder WITHIN the limits of the limit mate, all is well. BUT, when something pushes the cylinder beyond the constrains of the limit mate, errors and warnings should pop up.
To me, it is all about design intent. SWX does allow redundant mates (usually) to help capture design intent. To mate a bolt in a hole, you only need a concentric mate and a POINT on a surface. Yet, using two flat surfaces helps capture design intent (the bolt head is flat on the surface) AND it is allowed by SWX. A rather drastic example, I admit, but does illustrate my point.
Arlin (remove '351' from email to reply)