Now, when talking about roll cage design, we're talking about people's safety, so everything has to be considered, from the vehicle's gross loaded weight to the anticipated speeds of said vehicle, to the terrain that you intend to use it in.
For example, if your vehicle weighs 3000# and speeds around a track at 160mph, then you're going to want to ensure that if you roll at high speed and the high-speed momentum of that vehicle comes down at the peak of the A-pillar, then it needs to be as rigid as possible to disperse that weight. The momentum of that crash is likely to be greater than the momentum of a low-speed rollover for a 5000# vehicle, which is likely why there is this new requirement that the a-pillar bar can only have one bend in it.
In addition, the necessary diameter and wall thickness of selected tube for a 4500# FJ40 might not be appropriate for a 6000# FJ60 or 6500# FZJ80. The overall design has to change not only for the different design of the interior but also because the weight distribution for a wagon is different than that of a 40. Let's not even talk about how light those mini-truck things are
As for the roll-cage design with regards to terrain, I've seen some pretty open-top roll cage designs in open-top vehicles. If I owned a FJ40 with an open top and wheeled it like that, I'd be worried about a rock coming in from the top of the cage ( while upside down ) and knocking me in the head. Uwharrie has some pretty loose and jagged rock formations, and then obviously lots of trees + tree stumps, which all pose a head injury risk. If I had an open-top vehicle, I'd put more bars spanning the main side hoops to better protect the head. Better yet, skin the top with 14 gauge or better steel and put some speed holes in there. Take a look at this picture and then envision a pointy rock or tree stump sticking in through the head area of the cage and giving you a concussion ( pilfered from
http://metaltech4x4.com )
Real-world testing is the next best thing to CAD testing. 4xinnovations did a great series of tests that show the effectiveness of their cage. It would better represent a Land Cruiser weight range if they welded a bunch of steel plates to the floor.
One last thing to consider is that no rollover is perfect, nor is every rollover the same. Imagine a situation where you are in a rollover and instead of rolling down the mountain, you roll over twice and then into a huge oak tree. Is your roll cage designed to protect in event of a side-impact? Take a look at this video:
Overall, there are a TON of things to consider when improving the safety of your custom vehicle, not just your average rollover. I haven't decided if I want to put a fully-triangulated cage in my truck yet because in order for it to do what I want it to do, I'd need a bunch more tube than your average cage.