I looked through your lead-length article.  Pretty slick for just starting up.

I did a bunch of work on lead length issues for the 1299 standard.  I came to realize that inductance is defined only in a closed loop path, and actually, most of the formulas for inductance assume conductors of infinite length. 

There are a bunch of hairy equations to define the inductance of a short length of conductor.  It all comes down to the N*d(Phi)/dt of the closed loop defined by the transformer leads, transformer, and arrester.  Actually, with the "zero" length configuration, there is still some inductive effect caused by flux induced in this loop.  However, the guidance you give is about the best we can do. When I compared the simplified  L*dI/dt approach with the detailed flux-coupling analysis for a range of typical riser pole cable terminations, and the simple approach is "good enough" most of the time.