What I don't like in such articles is that they never give a clue how one can go from the bad category (obsessive passion) to the good (harmonious passion), and whether it's ever possible.
in my case, i think i have become "better", and what helped was a combination of understanding that i could burn out (when you're young you think that you are indestructible!) and decoupling my "own" projects from my "work" projects.
the decoupling is a difficult balancing act. an employer wants you to care, but wants to avoid conflict of interest. luckily in software development you can develop a personal project that is overtly about a completely different subject area (software for model trains say, rather than payroll systems) but which still improves the skills you use at work (via shared technology, concepts, etc). so while i tend to spend spare cycles on work that doesn't directly help my employer, i am also the person suggesting new approaches and solutions to problems at work - imho it can be win-win.
incidentally, i think the decoupling helps because it provides you with more control (see article). so that suggests that employers who want employees to be as involved as possible, but to avoid burnout, really need to relinquish control - to the extent that some kind of "dual ownership" exists (whatever that means - i am just exploring the idea here).