I need to read the article, and mull over the remaining disciplines, but I discount both Quality Engineering and Software Engineering from the list because neither area has sufficient reliance on advanced math (read: Calculus) to be comparable to EE, ME, CE, ChemEng, etc. that make up today's existing licensed areas.
Yes, I realize Quality Engineering relies on a ton of statistics theory that EE, ME, etc. don't, but I've found myself having to learn that on my own in order to effectively practice in my area of expertise. Thus, I see QE's core as a subset of the existing licensed areas, and their cores missing from QE.
Yes, I also know that Software Engineering can require mastery of Calculus, depending on if one is developing from scratch. In general, though, I've observed that the discipline includes little, if any, advanced math education for the majority of people that practice it, based on my experience.
WRT licensure based on experience, I'm all for it, provided the candidate either passes the same written exams OR sits for an oral exam with their state's examining board AND that board does more than just share war stories with the candidate (i.e.--asks some substantive questions that will demonstrate the candidate's ability to think on their feet and demonstrate mastery of the fundamentals).
I'll go along with substitution of experience for education as a prequalifier (most certifying bodies for other disciplines that I'm familiar with allow it, some in the converse direction, too), but feel strongly about the need to demonstrate the ability to pass exams.
Doubtful that oral exams would/could reemerge, but it's a thought, and one that makes things a little easier for the experienced but not formally educated practicing "engineer".