Draft

Since 'draft' is technically the pressure differential btw the flue gases and the CAZ due to reduced flue gas density from the heat, we use that as an indirect measure of mass flow. We say a venting system 'has a good draft' when in actuality, we mean it is flowing about 0.8 meters per second according to ASHRAE. That means at that pressure differential we get a certain velocity, which, when it encounters flow restrictions, reduces the actual flow compared to a short, straight, smooth conduit, we interpret that as 'draft'.

A sig. or smoke source can only demonstrate a draft hood is just as stupid as those who think it serves a safety function. It is a separation of noxious flue gases from the one dedicated exhaust conduit with no instructions, directions or legal requirements for those gases to find their way into said conduit. Air and flue gases are stupid and will go where a pressure gradient forces them and there is a pathway for movement.

If we take CAZ air and suck it up a stack, that makes room for more stuff, such as the byproducts of combustion. Those get entrained with the CAZ air, now referred to as 'combustion air', combined with some extra CAZ air that was standing around not doing anything or paying attention and it all heads for the exit. However, some dope put a hallway at the room exit with big, wide-open doors but crosses his fingers everyone will instinctively know, without signs, to know which way to go and no penalty for cheating and taking the side short cuts.

If we allow just the flue gases to exhaust, some will hang out along the walls like punk kids at school and damage the walls. However, if we dump a lot more people moving briskly down the hall, everyone tends to move along and there isn't as much graffiti or damage to the walls of the hallway.

It helps to refer to it as 'draft pressure' and 'mass flow' of flue gases as two entirely distinct, separate concepts.

Draft hoods suck figuratively but not literally. You would have to apply smoke 360 degrees around a typical draft hood to ensure a lack of spillage. Some spill switches recommend doing this test to determine at what point on the compass you mount the switch. One problem is, you need it at one point normally then another when the adjacent clothes dryer kicks in or the furnace with the leaky return ducts fires up. Or, replace it with a barometric damper that focuses a backdraft directly towards one spill switch and you have a modicum of safety enhancement.