I think that assumption was based on how most navaids work.

Oh, no doubt. One might be led to think that there is more protected
area on the other side (where you overshoot the navaid) than on the
side from which you're approaching. I vaguely recall seeing that was
true in the enoute segment in certain circumstances (don't hold me to
this), but was NOT true in the approach segment.

at what DME readout should you start your turn?

Exactly. And even if *I* could do it, can I expect a student to be as
proficient at it or understand the nuances as to when it should be
applied? An instrument pilot shouldn't have to be a TERPS expert to
fly safely.

I think it more prudent to hand a student a set of simple rules that
will always work, which is why I'm uncomfortable recommending leading
turns on non-GPS approaches. I think it's OK, but I can't *prove* it.

The GPS I fly with works all those numbers (OK, it doesn't know
the

With GPS, different story. I presume that following any lead provided
is both safe and legal.