Not that I've made one or played with one... but I did see the thing in action. They were shark diving in Cocos islands. Basically, if you removed the exhaust T and ran it up over your shoulder, you would still have a sealed system until the exhaust, so there would be no way you would be inhaling a higher CO2... you're still pulling fresh air from the tank. In effect, you're grabbing the bubbles as they leave the exhaust, and directing them up and behind you. Seems to me that, while yes, it probably would increase the exhalation resistance, if the flow was directed up, they want to go that way. So if designed right, the bubbles will simply flow up into the "catcher" then their buoyancy would carry them up behind you without breathing effort.
I do remember, theirs kinda hinged at the forward point - close to the valve, then as they would get air into it, the tube going along the tank would lift up at about a 30 degree angle. Again - since it's simply "bubble powered" it would not seem like an exhalation effort would increase to the point where you could not exhale into it. In a full upright position, if you had to blow the water down - that would cause a huge exhalation problem. It takes a good blow to get an air bubble to travel down even a foot... but again, if they just follow a flow of up and out - physics tells me it would not have the other issues mentioned.
Still... I just humm...