Tom is correct in every respect, though he doesn't need me to confirm
this. It is fairly easy to show that the Trask transformer is
electrically equivalent to the popular trifilar-wound 4:1 voltage balun
when wound on a single toroidal core. As such, it has no output (I
wound and measured one) into a fully unbalanced load, and of course it
has no choking action at all.

I do not know about binocular cores. It would seem the transformer
works somewhat into an unbalanced load when built with these, maybe due
to imperfect flux coupling between the two holes? I haven't measured
one. And neither has Trask himself. Until he produces a true transfer
function plot into a balanced and fully unbalanced load and a choking
impedance plot we are left to guess. To use return loss plots to infer
correct operation of a two-port network is, um, unusual.

73,
Glenn AC7ZN