I don't like that prospect of pulling the #1074 through
the harness loop(s), or the awkward loading on the
BWL that causes that.
This of course presumes a non vertical loading profile.
I would point out that in a normal end of line tie-in procedure, the rope is in direct contact with the harness (ie direct nylon-to-nylon contact).
Also, the loading profile to the harness of the middle person is not going to be horizontal (ie directed along the x axis).
The most likely scenario for a climber tying-in to the middle of a rope is for glacier crossings.
If the middle climber falls [down] into a crevasse - the loading on that fallen climbers harness will actually be mostly in the vertical (y-axis direction).
If there was a concern with loading profile - it is perfectly reasonable to tie-in with dual #1074 Bowlines.
Tying-in with dual #1074 Bowlines solves all issues with horizontal loading profiles.
But, as I have pointed out, the loading on the middle climbers harness is never going to be in a purely horizontal (x axis) direction.
As for efficiency, this tying in isn't a major portion of
time & effort or gear for the typical party. Having a
'bner spared for they joint as I said might even be
seen "as a feature, not a but"! Who goies out with
too little gear to do that?
Efficiency has many metrics.
From a technical viewpoint, if a tie-in procedure requires a carabiner, then it cant be as efficient as a system that does not.
In other words, if a the use of a carabiner is a condition to achieve security, how can it be as efficient as a system that does not require any extraneous equipment?
Put the question out to some alpine guides and see
what those-who-might-use say --maybe adding some
further insight to the pros/cons of both?!
That argument is what is know as 'social proof'.
In other words, it hinges on the general approval of others for it to be true.
One could ask some alpine Guides if they join their accessory cords with a Zeppelin bend.
The answer would almost assuredly be "no".
Most would answer that they use #1415 Double fishermans bend to form a round sling from EN564 accessory cord.
This doesn't mean that a Zeppelin bend is 'wrong'.
The same could be said of what knot do alpine Guides use to join 2 climbing ropes for a retrievable abseil.
Some would answer; "#1410 Offset overhand bend"
Others would answer differently.
...
Okay...
The IGKT is a place to come to discuss technical matters.
Tying-in to the middle of a climbing rope (for a 3rd person) is a technical matter.
The reality is... there is no 100% perfect answer.
It comes down to judgement and experience.
I had advanced the notional concept of using #1074 Bowline as one possible solution.
I have tendered arguments in its favour.
However, i do
not dictate its use!
It is an individuals choice. There is no law or legislative directly to force a climber to use a particular mid-rope tie-in method.
The default solution has been to use #1053 Butterfly with 2 locking carabiners.
Is it acceptable?
Answer = yes
If we frame the question a little differently... Is there an alternative method that does not rely on metal connectors and yet, is also biaxially loadable?
The answer is yes - #1074 Bowline.
In terms of loading profile - which seems to be the key argument - rarely would it be in the horizontal (x axis) direction.
IF there was a concern about horizontal loading,
dual #1074 Bowlines completely solves that concern.
In terms of seeking social proof by asking other Guides...I would say that the notional concept of using #1074 Bowline (single or dual) hasn't entered into mainstream consciousness.