About your question 2.

Today, NC method appears to be based on arbitrary bolt tension/stress and nothing more. In addition, it belongs to ASME III and people are hesitant because obviously we not apply the specific requirements of ASME III to process piping.

But the truth is that formula appeared in an work focused on B16.5 flanges.
Just google "3445600189702.pdf Rodabaugh" and you can find it.
The title is "EVALUATION OF THE BOLTING AND FLANGES OF ANSI B16.5 FLANGED JOINTS:- by Rodabaugh and Moore.

Mr. Everett C. Rodabaugh passed away at age 100 and had a huge contributions in Nuclear Codes and ASME B31 Piping, however his Fortran software FLANGE wasn't fully accepted as replacing Taylor-Forge method. In fact FLANGE corrected some (wrong) assumptions made in T-F method that are responsible for inaccuracies one can see when compare with FEA methods. BTW, when I tried to get a formal access to FLANGE executable, I faced it is still under Nuclear rules, even the Fortran listing can be found in one of Rodabaugh works.

NC-3658.3 formula is a result of FLANGE evaluation of B16.5 flanges concentrated in an envelope formula.

See also some words in
http://forums.coade.com/ubbthreads/ubbthreads.php?ubb=showflat&Number=53645

BR