Hi Enhaut
This is a very clear and straightforward approach, combining two well known, basic, knotting elements.
I like the solid way you are binding the crossing knot, making this component more stable and less prone to shape distortion, with this figure eight entanglement (a known issue of crossing knot based eyeknots).
As much as i love the herringbone look, at C-jump-B snapshot (a quite "catchy" descriptor
), i consent to the original post's collar placement, which needs to be located behind the figure eight, in order to be somehow restrained from being deformed at peak loads.
I guess we don't want the jumping collar to fold out into a malformed collar, or even a "ghost" collar, do we?
A couple of quick tests, showed that your nub is very pliable, with no problems of coming undone, even with the reverse figure eight based, tangle (more stable, but much tighter as expected).
I usually tag such knots, as non-PET, pseudo bowlines, because of the remnant, nipping knot structure.
It may not be scientifically quite correct, but many times i give in to my curiosity, and perform
knot war tests, for instance this knot versus its reverse, and draw some conclusions about their overall response.
Maybe Alan would be willing to further investigate the stability/jamming threshold of this very promising, EEL, TIB, eyeknot?