Hi Alan, i see that you have advanced from first stage complexity (Lee's link bowline) to second stage, employing additional tail maneuvers at the returning bight leg.
I can't tell which approach is neater, i like both, as a matter of fact i was thinking of inducing a round turn at the first returning stage of link bowline.
Surely such a move would embed an overhand in the system, but at least it would maintain the TIB property.
On the contrary, your round turn component around the nipping loop, is constructed in a way to stay clear of overhands!
Xarax claims that a round turn component, is almost equivalent to an overhand in terms of generated friction, it does not make use of a counter -measure mechanism to oppose to jamming, and he certainly has a point there.
In me view, it also has to do with the number of rope diameters feed through the round turn channel, in this specific case there are three.
If this overfeeding of round turn and nipping loop, (three lines in both) has an aftereffect to nub's pliability, i surely don't know.
If i separate the returning structure in five stages, the first three are identical to Scott's locked (TIB version), the distinction lies to the second turn which places the tail up through the collar and then back through the round turn (it is feasible to bury it even more by threading it for a fourth time down through the nipping loop!!!).
For me , I see no security/ stability issue coming out of this profile, but we have to also weigh the viewpoint of the wellgrounded experts in life critical applications.
Is there a quick test?