i think this is all consistent with all knotting/hitching and rigging. Not an isolated study of it's own.
If load force potential is too much for light sling and krab set; i place a half hitch prior to the sling/krab set.
Adjustable sling eyes are best with Standing Part on adjsutable end; instead of just wearing out 1 point on static eye end. Round slings are best rotated so that krab pull is also from different points.
A Round Turn Basket or Double Basket can give grip and strength. Also, Basket can be below stob w/ Center of Gravity on the other end or when pull from support pulley on sling/krab is downward into stob etc.
Linear use of sling is 100%, Basket is 200% and Choke/Girth are 80%. Mathematically; this comes out to a Basket being 2.5x as strong a position as a Girth/Choke or a Girth/Choke being 40% of Basket. The strength loss of the Girth/Choke is from the system bending it's own Standing Part; such as many hitches/bends/knots do; that gives them a ceiling of 80% tensile strength. The Basket offers 2 legs of support (in eye to eye sling or 4 in round sling if counting loop as 2 legs of support) where the others offer only 1(or 2 if counting loop as 2 seperate legs).
But, these figures are all assuming straight legs. Whereby drawing legs of basket together to place in krab places them at a leveraged angle/ not precisely inline to the initiating pull of the krab and rigging line. The legs of sling support run much more perpendicular to the initiating pull with a flat 'teepee'/ point of basket coming together; giving high leveraging of the device. This happens most extremely when sling barely reaches around host/load. Of course single linear or choke supports do not have this problem; but 2 supporting ones may; unless they are straight/ pairallell to each other.
This is why on loading docks they use spreading sticks at the top of a crate where teepee starts; so that the resulting leveraged pull inward doesn't crush said crate in Baskets and Double Linear hitchings as the usually come to point at hook. Just like a krab; most strength from hardware is down the back, solid spine; a krab is just a hook with a fancy mousing.
A Tensionless Hitch with turns on one anchor device, then secured to another seeks not to have the failing of bending it's own Standing Part. But here we have to factor in other 'worries' consistent to other riggings; now that bending the Standing Part is peeled back/off to reveal these other factors. If the bend around load/device is too tight; the inner part of the arc is compressed/ not pulling; and only the outer fibers really work; degrading the potential tensile strength. This will be determined by how much the hitching resists the bending/ therefore allowing leveraging. Then, that factor multiplied by the height/thickness at bend. Thus, flat webbing is used many times as it loses less strength; as it resists bending less; and has no dimension on the bend to multiply by. Thus both multiplying factors are as low as possible to lessen useable tensile strength by.
Another consideration that a Tensionless Hitch can't escape, that affects many other hitches/knots is the way the hitch device stretches consistently until the hitch/knot is formed. Then stretching reduces signifigantly. Thus, thinning the line area most right before hitch/knot. Almost like scribing a line across drywall; or making a notch in tree; all force finding this weakness/releif and breaking there. Especially when combined with increased forces of bending Standing Part in same area. i think this will be less noticeable in static lines; though they are stiffer/ resist bending more; so getting leveraged by that factor.
These are accepted numbers in industry; but can vary slightly; thus giving some variances in numbers. In fact a Choke/ Girth can be bent Backwards to place a 2/1 by Standing Part on restriction. Thus we wouldn't worry about the 20% loss of bending the Standing Part; and more about the 2/1 force of Standing Part. This reverse is also consistent with other rigging aspects. Whereby in bends in Zer0 degree range (2/1) we get more force pulling at end as input to the machine and picking up increased force at the bend. But in line bends near 180 degree range; we get more force return by inputting force at the bend and picking up return at the end(so is reverse strategy).
i think that consistent reports of 3 turn anchor to self to form noose etc. breaking the 80% barrier is from the force in the turns fortifying the Standing Part as it feeds into the hitch/knot and making more gradual/ not as sharp/sudden thininning and bending in this area. Notice if the host/load is given a Round Turn that the effect is lost; further re-affirming theory that the residual pressure in Bitters on Standing Part is needed to reinforce this problem area. The Round Turn on load/ host reducing the force in Bitters.
Many knots/ hitches are designed to pull perpendicular to the load (Running Bowline, Timber, Anchor etc.) to place Bitters inline with Standing; so are best preceded with a Half Hitch if load pulled is pairallell with line. This is why the ABOK knotting bible places such pulls in seperate chapters: 21)Hitches to Spar and Rail (Right-Angle Pull) and 22)Hitchings to Masts, Rigging and Cable. (Lengthwise Pull).
Fascinating study
; of consistent principals IMLHO.