Dang, look what eye missed..
Pulling straight is just your linear/inline therefore unleveraged pull (x1). Coming sideways/opposite of inline with your force; will give a leveraged angle of force on the pull/ per the resistance of bending of the line. Depending on the angle of that force on the rope that multiplier can be less than, equal to or greater than one. Almost a flat line but bent; gives multipliers greater than 1; multiplying your effort before it's output pulls on the target/load. At 120 spread at the bend; the multiplier will be one; so the line tension produced, pulling on the load will match the effort input. While more bent gives multipliers less than 1/ therefore speed gained; but power lost in trade going from input effort to the output force on target.
Per the resistance in bend; means that the line must be tight to resist the bend; like a wooden wrench could flex and lose potential leverage during application of force; the more the rope is tight/resists bending, the more the input effort is multiplied. This is trading distance/speed for power with these angles.
But, you are pulling against 2 points(target tree and anchor tree on either side of you); so your effort is divided by 2; then multiplied by the secant (1/cos) of the angle of deflection from inline to the target tree(inline being always the base/unaltered force). So, the increase multiplier of input effort to get higher output force on load is small; but effective. So, if you had a long pull; your inline pull c/would be better; because the bent line system only has so far a draw of distance to convert into power on target of power. Then you could take your increased tension/purchase and snatch it around the anchor tree possibly; this would be called sweating or swigging the line for more purchase(of line from the loaded side of anchor to the control/unloaded side of anchor) leaving the tension side of line tighter and/or the load moved.
So, the real bet is that if your inline(multiplier 1) input X it's faster speed/more distance over time is a greater quantity than the leveraged (multiplier greater than 1; therefore increased) input X it's slower speed is a greater quantity. In times that speed of adjustmeant, fast pull etc. can make the difference over force; we'd go with the linear pull over the side/leveraged pull to capitalize on the speed part of the inline/unmultiplied pull. But, of curse when the output effect on the work is raw power/not speed; then that faster inline pull's speed becomes a deficit; and we'd leave that behind and capitalize on the bent line strategy as Nature offered both to us. Anything; can be inverted or turned against 'ye to it's equal/opposite; with 1 simple change in this mechanic's game; that is what makes it amazing and dangerous to dance with!
i have come around an anchor tree with pull line in giant U tightening as you go; then sweat tight. Then turn that U into a giant muenter around tree with first turn/wrap over the Standing Part to target tree/ to resists and lock against it's pull(especially if going to sue to lower). Then slide muenter to side of target tree, then draw muenter as 2:1 pull that levarges across the Standing Part going to target tree. After Muenter slides back towards you as it tightens line; stop and push muenter back away to draw again. This gives a fair lock to fight a wandering load, a 2:1 leveraging to tighten; and can even lower(but note again first wrap of muenter goes over/resists Standing line's upward pull to target tree). i've used this leveraging contraption for very many things; very successfully.
This only works until each leg is 60deg; total spread is 120; then the effort input is equal to line tension; folded more close; the line tension is not greater than your input, but less. Before 120 degrees; we'd rather input effort at a bend and then take the leveraged increase out from an end (pulling on target tree). But; more closed than 120; you'd rather pull at an end and have the load at the bend. For, we'd rather be doing a 2:1 type rope shape on the load (with load at bend/on pulley). For, our pulling across on a rope; is really pulling backwards on a wide angle 2:1 pulley system. So; where it would take a bunch of effort to lift a block high(because the block has leveraged position over our position) to a straight line across 2 points by pulling from an end; we just take the same model and we input effort where the block would pull across, and place the block on the end of the line. Now, the block doesn't have leverage over us (because we can move it with less effort at our angle on rope than at a straight/inline/unaltered by multipliers type of input force on the line); we have reversed positions in the same formulae/model with the block. Then we extend that same model to pulling our tree over
The same science from a different perspective/to give a different view of same workings; is a Running Bowline. When the teepee is sharp the Bowline eye pulls at the Standing; but at a more inline angle. But, if we cinch it up tighter; drawing it in towards the load; the eye of the Bowline pulls more abruptly/perpendicularly/not inline across the single support of the Standing Part(which is under tension/resists bending/therefore can be leveraged).