i see relief as Up and Left(?).
i think i'd go with longer than firewood size, depending on safety of cutting height though. i think that if we scribe a wall board and it breaks clean that is from the leverage working from both sides of the scribe, to the only weakness of the scribe. i think if the wallboard was all the same thickness as wallboard left after the scribe; there wouldn't be all strength from both sides funneling into find single, small area of relief. So if whole board was as thick as scribed part, it wouldn't break so easily; for it wouldn't be weakspot, nor have stiffer pieces pushing into that weakness! So, i think we'd want more leverage on stump side (for more push from that side)so a little farther from there than firewood length. This would also put less flex at spar above cut(that would be usually thinner/ more flexible and have more force on it)than the lower part of spar you were placing the more leverage on); to hopefully place flex at cut; with stiff power coming from both sides.
i'd make narrow face up, then widen left side of the face to offer path of least resistance to that side on the 'scribe' in spar, and kerf right side of face. Then cut upwards; trying to close right side, so that it and supporting tree pushed left. While the left side of the face was calling it that way as only point of relief (in tandem with more relief from supporting tree on that side) and also as left corner of face was pulling it left too. Once the joint is loaded(as spar fails and begins to think about moving), we can work push and pull forces to equal force in hung spar). Here we work them in tandem; not one at a time(usually we work them one at a time with full face open , then full face closed, instead of this mix of 1 side open and one side close). Force loaded into open area pulling, while force loaded into close area pushes as the workable, mechanical commands to give system with saw.
Also, we work the 2 forces in opposite directions; with push apart and pull together commands; if they were inline; they would fight and lessen each other; spread apart they give tourque(as the forward part of a wheel moves down as the opposite side moves up); more efficient use of the same amount of useable force (whose capacity is to only match the leveraged force of spar).
i think this folding would be easier to trigger to failure, to self work the rest of the way; as you were out of the batter's box. If i pulled with rope it would be straight ahead more, to make tree 'heavier' in that direction or to left s-lightly; and force more flexing of the hinge earlier without cutter in batter's box; and trusting push in face and pull in hinge forces to correctly guide it.
Pulling to steer away from front of hinge unloads / replaces the automatically responding hinge power(with your force); pulling inline with hinge loads it, and empowers hinge's corrective forces(multiplier of your force) more IMLHO.
Or, something like that