I think I am beginning to grasp what you are getting at.
I can only help if I know where you go wrong.
But this is really darn simple. (Precise dressing
maybe will reveal challenges, but conceptually
the knot structure should be perspicuous.)
THESE '|' (bars) are material making a turn! [artist's rendering]
My flattened "Z" structure:
v----: tail
________________________________________
---------------------------------------------------------| < -- bight tip
|_________________________________________________________ _ _ _ main line
^ fold of bight back from mainline.
>>> If I am correct, we have dispensed with placing
>>> the line around something solid, making 20 -30 outward twists
Well, did I say any of that? (no).
One might have a *thin* at the "fold...back" point,
which will be where the bight tip ultimately tucks out
though to become the eye.
So, one needs adequate length to
twist main line & bight rightwards (per above sketch),
and then come back with bight overwrapping these
twists (which should be easier than what I just watched
for the Alberto, where one brings back a fiddly little single line!)
until reaching the fold-back point now reduced to small aperture
to tuck out the overwrapping bight to become the eye.
Generally, a "TWIST" will be a helix w/much greater angle
than the ensuing "OVERWRAPS" --though these in being
done with a bight (so, twin lines) won't be so numerous
as the single-line wraps in the
Bimini. AND one is hoping
for some good "imparting twist" action of the eye legs
going into the counterwraps so one might not need so
many, to effect & hold the mainline's being twisted/curved
in the knot's core.
YMMV ??
As for setting, in rope at least I find myself just working
--as presented orientation above-- right-to-left tightening
the overwraps, working around the structure getting the
material set well and excess moves leftwards into eye.
Then perhaps some tensioning of eye vs. knot as a first
step is good?
Damn typical of fishing-knots presentation,
but one simply canNOT see what the final knot
is supposed to be --you have to gander at an
ambiguous squiggle, just hoping that your tying
has got you to the right state (and, frankly, many
*expert* advisors/teachers might not know what
that state is --or what possible variety of set states
exist-- themselves!).
(-;