The most obvious one: One of the free ends of the bend is still free when used in the loop form, and the other free end of the bend is employed in making the loop.
I can't imagine leaving one of the standing parts of the Hunter's Bend free and still calling it a Hunter's Loop.
But this is
exactly what is done for the so-called "Butterfly Bend" (aka "Strait B.")
in going from IT to the better-known LK (but, yes, the origin came the other way 'round)!
And as previously noted, there are some interesting other loopknots that might seem
better suited to the moniker, as found e.g. for Rosendahl's Zep. bend, where essentially
one of the bend's standing parts is conceptually (and then, acutally) doubled to become
the loopknot's eye (and so preserves the exact sort of loading profile as the bend!),
and one finishes the knot appropriately. --less material efficiently than just making
the eye w/end of first-formed component and re-entering qua 2nd SPart of bend,
but, again, compelling from the aspect of loading, at least as a claimant on the moniker.
BTW, Ashley's
original work contained what is essentially a variation of SmitHunter's
bend--#1453. It is 1425a but where the rope leads to the other side of the SParts
in flowing into the respective
collars around them. It is less prone to jamming.
There is also a simple version of SmitHunter's bend where the ends sort of
embracein passing each other, and this knot better resists jamming, as material of each end
gets pressed a little into the collars and keeps them open. (So, ideally, 1425a would've
been (and there's enough page space) "1453a"!)
(Incidentally, the Bowline is often given as the loopknot corresponding to the
Sheet Bend; but as the bend is asymmetric, the Bwl corresponds to taking
one SB SPart qua LK SPart, and taking the other (the SB's bight half) one
is led to Ashley's #1058, a Fig-8-based LK !)
--dl*
====