I tried to illustrate there different rigging scenarios. The Paint image is starting with a static load set-up.
Tree one is a Lowering Device LD, a low friction block, Force 1 F1 weight of the load, F2 equal and opposite force to hold load stationary. There will be a bit of friction at the block, maybe 5-10% loss of efficiency, so F3 will be about 1.9 times of the load, so for simplicity call it double the load on the rigging point. F3~200 pounds.
Tree two is a lowering device with a NC rig, which will have more friction, and will be much less predictable friction. The counteracting force F2 will be less to keep the load stationary (I don't have a good number here, can we say 30% loss of efficiency due to friction, friction eats 30% of the load, so F1= 100 pounds, F2= 70 pounds (100-30), and F3= 170 pounds.
Tree 3 has enough friction at a LD or NOT-Redirected trunk wrap to hold the load stationary, so the load generates F1 which is 100 pounds.
Natural crotch rigging with a redirect, as far as I can tell will generate less force on the anchor than a low friction block for STATIC loads.
A LD or non-redirected trunk wrap will have the lowest force on the trees.
This is not going into compressional loading of the stems versus side loading.
When the scenario changes to dynamic loading, the ability to predictably absorb the energy of the falling load MAY be greatest with a low friction block system and a lot more rope in the system, as with Tree 1. Add into the equation the nearly equal counter-force F2 to F1.
In Tree 2, F2, the counter-force, will be less than F1 because of friction at the crotch, and the ability to effectively cushion the load F1 may be minorly or drastically reduced.
In Tree 3, there is no counter-force, but little rope (and stretch) in the system as compared to Tree 1 or 2.
My take is that in different scenarios, you will get different max forces on the rigging point F3.
Rope stretch, space to let it run, and ability to predict the friction of a LD and/ or friction wrap make it variable, with no one-scenario-fits-all option.
Thoughts?