Hi Ychan, it's good that you have included the interlinked configuration as well, providing a compehensive picture of these structures.
The "Beta", butterfly-ish variation, although it looks more compact, in Mobius, crossed-line style, i believe it blocks the access to the figure-9 collar, therefore, i also agree with your preference to the other one.
Afterall, what's the point of inducing additional complexity to 1053, which would probably render the untying situation, more challenging than the parent knot?
On the contrary, at the february structure, the 9-collar is completely accessible, with no direct linkage to SParts, and moreover, it operates as a blocking component against the distancing developed between the two non-interlinked overhands of false butterfly, holding them close to one other.
To see what i mean, try to form the figure nine, this time from the right side of the knot, incorporating an S-half hitch.
You will rather uncover a knot, not so stable, as the one with the Z-half hitch, that you have very smartly intergrated from the left side of the original knot.
One other interesting aspect is, that if the double splayed loop, bowline version is obtained from the original, inline, source knot, this figure-9 form, is transformed to Alpineer's nipping structure.