On the belt engagement, the belt needs to be fully tensioned to get proper belt engagement with the pulley/sprockets and, even then, the static tension is no where near what the belt experiences along the top run in use. More on belt tension in my 24 Nov 2010 blog entry on the subject. As someone else noted, there sure seems to be a lot of slack in that belt. Most of the belts I've worked on are very tight even with the eccentric in the most re-ward position.

As for alignment, you'll only know if you've got the alignment dialed-in when you tension the belt and can successfully rotate and counter-rotate the cranks in the work stand without the belt walking off the pulley/sprockets.

Finally, as for having the rings reversed with the flange on the inboard side... assuming the belt can be used on your tandem (re: length issue) both of you should keep a close eye on the belt during your initial rides. If either of you see the belt walking off your respective pulley/sprocket, you'll have your answer. In theory -- and with both the sprockets aligned -- the belt 'should' track quite well even without any flange(s). Taking it a step further, if your cranks, bottom brackets and boom tube have zero deflection under load.. there shouldn't be any issues. Therefore, the question becomes, how much if any elasticity is there in the bottom end of your Calfee as its configured? I would expect it to be fairly low (a lot lower than steel, Ti or aluminum), but I'm not sure if it's zero or just how much tolerance the belt drive will have for any rider-induced sync drive-side deflection.

A friend of ours who did a lot of independent work on tandem sync-drive's using the pre-Carbon Drive synchronous belts from Gates and who beta tested various configurations on their Seven 007 titanium tandem theorized that early testing on steel and alloy tandem frames that are prone to some bottom end sync-drive side deflection would have revealed the need for at least one of the flanges having an outboard flange to keep the belt from walking under heavy pedal loading, but never bothered to test any sprockets that didn't have flanges on both the inboard and outboard edges of his custom machined pulley/sprockets.