Mark's knot has the double-overhand dressed differently,
my knot has the double-overhand dressed as you would in a double fisherman's,
and it chokes the two strands of rope before passing through the single-overhand,
which appears to me at least to choke the ropes a lot better,
preventing the knot from spreading as much.
And this was my thinking, too, having developed solutions
the this rope problem like yours. BUT, I've slowly come to
see that Mark's later added tuck of the tail works well as a
*lock* on that tail pulling out --drawn out by the forces
against it choking (just as it would were it tied off in an
overhand knot around the other tail.
The full turn of the choking strand does as you say, makes
a stronger choke. The simplest way to implement this is
in what I've named the
"offset 9-Oh", where the first
term is an adjective indicating the general nature of what
one desires in an abseil-ropes-joining knot --being offset
from the axis of tension. (No, it is NOT "flat" --lousy term,
though popular (but then so is "half a dbl.fish." where "strangle
knot" should be)).
I think that this gets you to an image :
https://igkt.net/sm/index.php?action=dlattach;topic=5555.0;attach=20150;imageand this to the related thread:
https://igkt.net/sm/index.php?topic=5555.0Franz Baochmann & I think also Heinz Prohaska have suggested
using an
offset grapevine bend; this also gets that
strangle
knot choking, but need not top it with another full
stranglewhen a mere
overhand works.
I've come around to recognizing that the --to coin a name--
"EDK-backed EDK" which Needlesports promoted (about
a decade & half? ago), though clumsy looking, is a simple
and effective way to guard against mistakes, as I think
it can suffer various imperfections of dressing/tying and
still hold. --in contrast to the more clever knots that
require making the right tying decisions at various points.
--dl*
====