rubbinz:
Ideally, the tie in would be directly at the suspension mounting points so that no stress transferred along the frame and almost all of it ran linearly along straight spans from node to node. But things are rarely ideal (re. my rear suspension post in CoG). But as long as it's relatively close, or tied in on both sides of the suspension mounts, or otherwise in some way limiting the leverage of the suspension against the part fixed by the cage, it should be fine for any reasonable time span. Problems really start to mount with the "typical" family style cages in K5s and the like where the cage is well made and rigid while also tied to the frame at the bottom of the foot well. Already a high stress area (where the frame turns up), the leverage of the suspension bending against that point without allowing the flex to "travel" will multiply the fatigue at the bend. Will it fail right away, certainly not. Just as an example, what may have otherwise lasted many years and 100k miles of moderate wheeling and daily use may now last only 2-3 years or 20k, or more, or less…
2dogs:
Ok, I get it now about the cage. Yes, I agree that you do not want to take a chance on damaging the cage by trying to co-opt it for another purpose (like stopping flex). However, properly designed and triangulated structure utilizing "nodes" and straight spans can handle those loads just fine without damage. But more aesthetically pleasing designs are often much less rigid in general and weaker when called on. However, people will still make those compromises…
Your also right about the flexing, to an extend. Problem is, people see the end-to-end flexing in the stock ladder frames and say "Oh my! Gotta do something about that!" And they tie in the cage, or scab on some plate or partial boxing, and they see "reduced flex", and think they have succeeded. What they don't realize is that small movements are still taking place and are localized to a very narrow area, with the net effect being reduced time to failure at that point.
