Great discussion!
My rules of thumb are similar. But mainly i watch 2 things specifically to determine exactly how hard to apply these standards.
1) is to undermine beyond the Center of Gravity well enough to a)want to fall in chosen direction + b)flex the fiber of hinge to do so Also asses how high CG is from proposed hinge pivot height.
2)Bring the back of the back of the hinge to the approximate widest part of the tree; especially on the control/anti-lean side for best leverage against sidelean or side pushes during fall (from other obstacles in path).
No. 1 is generally preset for you; lean away from hinge; hopefully fairly in falling direction. But covers the short/stubby extreme example by SmokeChaser too. If you can't bring the hinge pivot back far enough to accomplish this; you must either push or pull.
No. 2 Takes into consideration sidelean control needs and non-perfectly round stumps, rotted areas etc. i always face tree; then inspect face for decay and dryness. Decay is bad/bad. Dry/dead/ frozen wood can still push; but is less helpful for tension/pull; being less elastic. Both push and pull is needed in hinge/face machine. So i assess where tension and compression wood is needed; and if the face shows that type of condition possible in those specific positions. And dead and decayed are 2 different categories.
i believe in push (mostly by wedge) and pull (mostly by rope) to force a stronger hinge; just relieve these added forces when tree starts to move. Forces at FirstFolding(i define this time period as done when the tree will travel on it's own) determine hinge strength; more force forces hinge stronger. Hinge strength just being a reflection of the forces on it at this point in time; so we 'fake it out', to 'think' it has a bigger tree to steer! Additional forces after FirstFolding tax the previously set strength of the hinge; so essentially weaken it; so we relieve them. This is one thing a wedge works for us automatically; when the tree lifts, the wedge stops pushing.
i think the slanted BackCut is a no-no; in that the 'back stop' is weaker in case of sit back. If the tree sits back against a slanted back stop; the immense leveraged force goes more across the column of the stump/ rather than served down into the strength of the column. Also, there is weaker backstop to push off with wedge if needed for the same reason; and the wedge is pushing more straight ahead to bring tree straight ahead. But in a straight backcut with wedge pounded in, the wedge lifts, to serve forward. So the former is more of a linear applied force, whereby the latter is more leveraged arc; taking a longer route to achieve same result/direction factor. Another consideration in wedging is that the farther back the hinge; the less leveraged distance wedge has from it's lift position to the pivot of the hinge.
i might have slanted the backcut on one of them trees; to not cut into another tree behind it. and still have hinge leverage/height where i wanted; if it was safer like that.
In good wood/up to the task, i prefer to let the hinge steer mostly; and apply all additional / hinge strength increasing forces(rope and wedge) to the gunned face/ center. If pulling with line or pushing with wedge to steer; then we are unloading the hinge from it's task some. i'd rather force the hinge stronger, then taper the hinge to steer; so that my exerted additional forces go through another multiplier (hinge); instead of unloading hinge from imbalance, and replacing it with my effort; allow the mechanics to work for you.
Works great in-tree too; especially if you kerf dutch the sidelean or push from side(not full face) to work in concert with the extra pull of tapered hinge. Then, we have a pull and push in same direction and rotational, with counter pull/pushes dropped out of the equation (when kerf closes and tapered doesn't).