Although I have driven a couple of EarthRoamers over 80,000 miles, quite a bit of it very rough terrain (e.e., raw lava fields in Eastern Oregon), I have no video. But I do know you have your hypothesized geometry wrong: the ER house is bolted directly to a couple of lateral extensions of the frame in the fwd two corners, and has a pivot in the back.
that's something of a relief to me. My belief is that rigidly fixing house to chassis (with torque free mount) behind the rear axle will in some circumstances create more roll of house and potentially adversely effect handling.
I hope this isn't "teaching grandmother to suck eggs". I haven't been at this long but I suspect that describing house movement relative to chassis and not describing both relative to the road may be misleading. "Torque Free" is catchy but only considers house sub frame movement relative to chassis twist not the equally important chassis roll and the dynamics of the whole system.
Maybe just me but I find it helps to consider everything relative to the road. The most "fixed" reference object. I prefer a mythical fixed, flat, road with bumps where I need them to change the equilibrium. It also helps to consider the chassis twist due to both front axle and rear axle roll. If the axles both roll to the same degree in the same direction at static equilibrium chassis twist will be approximately zero and chassis roll will approximate to axle roll. If the axles roll in opposing directions then chassis twist (relative to the road) at some point between the axles will be zero but chassis twist at the chassis extremes approximates to the sum of axle rolls allowing for some probable adjustments for suspension spring rates and chassis stiffness. I'll leave you work out the +- signs for clockwise and anticlockwise and contemplate the absolute movement in mm for any point in the chassis.
One hypothesis would suggest that the best place along the length of the chassis for a fixed body mount is where minimum body roll has to occur to accommodate chassis twist. Assuming same front and rear suspension spring rates that point is half way between the axles. But before we get too carried away, don't forget that the chassis also bends. The temptation is then to strengthen the chassis or spread the load. And if we do that it may lead to a stiffer chassis and the need to soften the suspension springs. And ....... etc.
Damping is interesting. Its a dynamic thing and some of us may not make it past the static aspects of the design. I don't know how others do it but I tend to think in terms of relative mass, inertia and momentum. The effect of the distance between house centre of gravity and house roll centre as well as the moment of inertia of the house and the distance between house centre of gravity and chassis roll centre. There's also the elasticity of mountings (even fixed mounts) and absorption of energy. And hopefully there aren't too many resonances.
Back to where the body fixed mount is for a 3 point mount. If its at the extreme rear and the axles roll in opposite directions the house will roll. At some speeds the momentum of the house roll will in turn attempt to further roll the rear of the chassis, which must be accommodated in the suspension springs.
Now for the messy bit. Designed into the suspension is "roll steer". Without it we would have oversteer whereas understeer is usually perceived as more desirable. Its part of creating predictable stable cornering. The axle movement where its attached to the leaf springs is not vertical, its about an arc centred on the front spring mount. Depends on the suspension geometry but when the chassis rolls relative to the rear axle one rear wheel moves forward a bit and the other back a bit which introduces a small but important steering effect. There is a similar effect on the front suspension where the axle roll is reduced. Front suspension geometry has a similar roll steering effect plus the effect of the steering coupling itself - the steering box is likely fixed at a different centre to the front spring mounts so for a fixed steering wheel position the steering movement at the wheels changes as the axle moves up/down and/or the chassis rolls relative to the axle.
The limits for the FG suspension appear to be quite wide, otherwise we'd hear a few more war stories. However, get it all wrong and we may proceed down the road meandering (potentially uncontrollably) from side to side with a driver tendancy to attempt to correct the yawing by turning the steering wheel.
Now try to picture it while driving round a bend.
To take it to completion, which I haven't, there's a need to build a picture of all movements from road (the most fixed reference) through all the mechanical transitions to house and back at all speeds considering the strength of all components plus the impact on ride and handling. Not a trivial exercise. At its worst there may be a point where it all becomes uncontrollably unstable.
For now I'm content that having the fixed mounts of a 3 point mount behind the rear axle is fundamentally unsound. They belong somewhere between the axles, preferably to achieve acceptable house and total roll relative to the road in most circumstances. I should add that this is only one small aspect of the total "body mount solution".
