Robert,
In the end, it may simply be, if the belt is as wide and thick as it is supposed to be and the motor/JS pulley is the diameter it is suppose to be, using a multi-ply, rubber-impregnated, canvas belt will work just fine. I mean, it IS the replacement evolution for leather.
This belt type has an inheirent 4-5% stretch, at the end of which it will not stretch more unless overloaded for it's construction (got that from Western Belting in Colorado). With the long-belt configuration you, Dan, and we have, I'm guessing the stretch would have to be removed by resplicing, though it could be removed by lengthening the distance between motor and drive pulley through shim removal, but that is self defeating to some extent.
It is not that once the 4-5% stretch is gone the belt has no give, though. It does, but the function becomes give and return to same length. I've got no idea how that works with the limitations of idler arm travel. Obviously, the wider the belt, the thicker the belt, and the shorter the belt (there is such a thing as too-short in the world of belt-types), the less temporary stretch for the force applied.
On the rubber impregnated canvas (my guess at it)... the rubber may avoid messing with leather belt dressing and cleaning. I've not read anything concrete about this, but little bits of discussion here and there makes me think it is part and parcel of running leather belts. Bruce may know more about this from his historical machinery interest.
To see a range of belt-types, search McMaster-Carr and then search "flat belts". There are single, double, double-with-woven-between, and others. I thought seriously about double-leather-with-woven-between... There are motor/JS pulley diameter limitations due to thickness. They are expensive. If they take a permanent stretch, shortening is onerous. They may need dressing/cleaning to function properly. Those are the reasons I am looking at other belt-types.
Mike