Still in "similar to rant" mode,
I'll toss out some general remarks about tie-in knots.
Knotting enthusiasts might relish a plethora of knots
to explore & compare ... , but most In-the-Wild
users--perhaps especially those in life-critical endeavors--
will want few(er) choices and commonality/standards;
they want to quickly use the "right" knot which will be
one that others know to expect and simply *know*
(and not DLehman's latest favorite flight of fancy!).
[ :: FYI, I've been pressing to illustrate & make presentation
firstly (ca. Oct'17(!)) of tie-in knots history
(the original, the changes, the mused-along-the-way),
with of course a look at some "bowlines" among others;
and currently --the first effort stalled-- it's now a matter
of what was first one set and then that expanded to now three sets
of "bowlinesque" knots --which is sometimes overwhelming
ME ("Do I have this, already?!" I keep wondering.)
E.g., consider (this isn't in but is somewhat like knots
in one set, in having an *eye collar*) loading the Angler's
/Perfection loop on its tail --which makes a nipping turn,
thus ... *bowline* (for ME)-- :: there are many such
knots that get some stability of orientation via an eye
collar, and one set in which one can swap the eye collar
with the eye (i.e., retract the eye through its collar and
extend that out through a now shrunken-to-collar eye
to be the new eye), and also can load the tail (of both
prior & subsquent such "swapped" knots).
For my sanity/clarity, I had to sketch out to see that
one *seed* knot could be tail-loaded (okay) or else
eye/eye-collar swapped for decent new knots; and
each of these getting the other's treatment (i.e.,
load tail of swapped knot & swap the tail-loaded one)
produced my "Doubly HAH!".
But reversing the simple tail's-turn crossing w/itself
(which made a fig.8-based like knot, just like my above
"seed" one BUT, unlike that, not TIB), though also
producing decent knots either tail-loaded or "swapped",
when going the further step of taking the other's
operation ... got ITSELF (and not a new corresponding
knot to the Doubly HAH!).
.:. All these structures had both ends make a turn
on entry (and one end would do more, the other
just exit into eye leg), and the latter case above
the turns were same handedness/chiralty, so that
led to the compounded operations returning to original.
(or so I think).
]
One can see knots from aspects of included structures.
E.g., it should be a simple matter to see a bowline and
then figure a double/round-turn bowline from it; to see
the basic 4-wraps Prusik & ProhGrip friction hitches and
figure out that 5-/6-/n-wrap versions are possible, and
sometimes beneficial!
To see a Yosemite tail wrap,
to see an "end-bound" wrap,
of that overhand-forming tail path Alan Lee included,
and, and the "Janus" collaring of an eye leg,
and so on ...
to producing *new* knots.
For tie-in knots, one should first get a good appraisal
of the materials & tyer needs. E.g., "ring-loading" : is it
really an issue? If one is tying so small an eye as is
shown in the OP's presentation, no; but if one has a large
eye or plans to use the eye qua *ring* for belaying
--a practice that has been shown (in some guide book,
even)--, then ... YES.
How about TIB? --hardly needed for usual tie-ins; but
might be immensely helpful in other situations, no end
in sight.
On overall structure : consider I think it's DAV's (German
climbing org.) recommendation for using the
bowline on
a bight --but reeved, not TIB--, which gives twin eyes
for tie-in. This had a perhaps advantage of being gentler
to the harness?! (2 strands vs 1 beating into webbing)
And to come completely untied, the tail must go a long
way, during which its flopping about should be obvious.
.:. So, one could make twin-eyes a design criterion
for selecting a tie-in knot.
And so I've just been revisiting the eye knots derivable
from the
overhand & fig.8 eye knots where the tail
reeves back into the base
as though to make an *offset*
end-2-end knot to see about putting in a 2nd eye
--sometimes with a simple tuck of tail through an ample
entanglement. (But on this simple 2nd-eye finish, I do
worry that if the knot were used for belaying there'd
be confusion about selecting a single eye when both
should be used.)
And, for that matter, I've long ago figured a better finish
for the DAV knot than a pure BoaB, which has the sort
of "overhand" tail finish of Alan Lee's-locked knot (and
can be, not that it matters (or is easily seen), TIB!).
Why tie the BoaB whose tail finishes rather lamely
for the purpose?!
Alternatively, one might seek a knotted structure that
is built up of individually decent components, expecting
redundancy to play the key safety role, and finding some
relief from more complex structures in a combination of
simple ones. Tying off whatever eye knot with a
strangleknot is an old one.
(To some extent, e.g.
Lee's Link Bowline is a redundant
variation of the "Myrtle" --which is possibly what remains
should Lee's final tuck come out (though one must ask
how that could be and in what state is the rest of the knot
then!). As can be seen of some other complex, extended
knots. )
One could envision and eye knot feeding a hitch where
the hitch's tail made a stopper before returning into the
eye knot to finish it :: even if the eye ... untied, the stopper
could hardly walk off the line and so should lock even a
loose hitch (!) --sort of a Rube Goldbery tie-in

. But
beyond the amusement, one might question the effect(s)
of any looseness in the run of material before a back-up
came to the rescue --think "friction & heat".
Re this last consideration (material movement),
I wonder at some eye knot such as the
mirrored bowlinethat should avoid being jammed tight, allowing some
relaxation of the structure to be available on a further
loading to compress and absorb force --whereas e.g.
a
fig.8 eye knot would jam hard and hold its compression--,
so that on a series of drops (FF1, say) one might see
a more rapidly increasing peak impact force for the F8
vs. the Mrd'BWL!? But maybe they break at the same
number of drops?! --with the F8 looking "stronger"
per higher force, but then consider that one's system
must endure those higher forces. And could this
result happen with a difference in rupture forces
greater than shown in slow-pull testing (presuming
that there is a difference, in the F8's favor),
which we might believe must've resulted from some
damage from the repeated compressions of the BWL
not felt by the F8?!
(I've NO idea of whether such is the case.)
(To which I must add that I wonder at the tensions
in the respective eye legs of the mirrored bowline,
in that the outgoing eye leg has two nipping turns
feeding it --and these are what will compress--
whereas the returning eye leg runs straight
through these turns. One can swap the lower
turn into the returning eye leg, for balance!?)
((ACK, now a new stream of knotting fancy is flowing ... !))
.:. Just more aspects to consider.
--dl*
====