When considering doing single channel MQL, whether through turret or spindle, the coolant channel passages will have an effect on how well it will work. Straight passages with no turns, cavities, or restrictions are best. Three hundred years ago, Newton observed that objects have a natural tendency to keep on doing what they are doing. More technically, an object in motion continues in motion with the same speed and in the same direction unless acted upon by an external force.
This applies to the aerosol droplets moving through the coolant channel as well. Bends and changes in diameter of the passage, especially when they are abrupt, cause some of droplets to hit the side of the coolant passage. The sharper the turn, the more that will hit.
Similarly, restrictions will cause more particles to hit the sides as they squeeze together to make it through. When the particle hits the wall of the channel, it wets out on the sides. The more particles cling to the wall of the coolant passage, the less consistent the MQL output will be.
When the air is flowing, this build up on the walls continues until enough particles group together that the airflow can rip them off. When this happens, a large globule works down to the tool and then comes out in larger drops or runs. You will always get some of this, but it gets progressively worse as more fluid sticks to the passage walls. This is, in part, what causes the sputtering and spitting when the system is running.
Of course, not all of the fluid will be stripped from the walls by the airstream. Gravity will eventually take care of this. So when the tool is changed, or when the machine is sitting, the oil will eventually run out. This is inherent in a single channel system. You can minimize the spitting and sputtering, but you can never fully eliminate it.